100 Most Recent Reviews

• iwxNZwcnIb4GC9sM8Hv6kw==October 5, 2026summer 2026

  Video Game Design and Programming

  Fun class. Learned alot. It takes a good bit of time to do well.

  Rating: 4 / 5Difficulty: 3 / 5Workload: 15 hours / week

• hhZi/tShOVopIm/TstBpcQ==September 7, 2026summer 2026

  Knowledge-Based AI

  For context, I took this along with Computer Networks and ended up with a borderline A. The content is relatively interesting as it helps think about how modern day agents follow the heuristics provided in the course which were designed over decades ago. I think the class would be easier if taken during the Fall or Spring since during the summer, there was something due every single week, not giving me a life outside of homework and my job.

  However, I think the ARC AGI portion of the class is very tedious and rather annoying. It incentivizes students to use brute force approaches if we are not able to conjure up a solution using the heuristics from the lectures. This is what makes the class tedious as towards the later milestones, all you can think about is getting the course over with. In addition, every report which goes along with every assignment (classic Dr. Joyner class) feels tedious as well, since I felt like I was repeating myself from one milestone to the next. And the TAs seem to be very particular about the analysis you do in your report, so follow the rubric closely and always over explain.

  For my semester, the original labs were supposed to be based on some KBAI agents developed by PhD students where we interacted with it and answer questions in report format. After the first two labs, either due to technical difficulty or push back from students regarding the usefulness of the labs, they made the original labs into participation grades and redid 3 labs to actually implement concepts from the lecture in the coding format. Of course, reports went along with the labs as well where we explain the development process and connect it to lecture material.

  TL;DR Lecture material is rather interesting as it helps understand heuristics behind modern agentic design. However, assignments are rather tedious due to repetition and reports. Good class to pair with an easier one as I did with Computer Networks.

  Rating: 3 / 5Difficulty: 3 / 5Workload: 17 hours / week

• Leq5tv8IK8tghRyJLzn6eA==September 4, 2026spring 2026

  Data Analytics in Business

  I had read the reviews of the course before enrolling, so I kinda knew what to expect. This meant that I didn't watch any of the classes, but jumped straight to the homework (which felt a bit outdated). The only challenge in all of them was to make proper API calls, as the code development itself was very straightforward. What I really liked about the course was the final project. I did all the previous ones by myself, but in this class you're forced to create something with teammates, which taught me a lot. I really enjoyed our final project and all the projects I reviewed.

  Rating: 3 / 5Difficulty: 2 / 5Workload: 8 hours / week

• l6Z2LEWR4pUHmqtA4hR5Kg==September 3, 2026spring 2026

  Software Architecture and Design

  This review and course summary was taken from the foreword of my SAD course study packet, available on Apple Books (https://books.apple.com/us/book/software-architecture-design/id6799493659) and Amazon (https://www.amazon.com/gp/product/B0HDGY48S2?storeType=ebooks&sr=8-1):

  These course notes represent my efforts to edit, summarize, and catalog the lectures for the Software Architecture and Design (CS-6310) course at Georgia Tech, taken in Spring 2026 as part of Tech’s innovative, fully online master of science in computer science (OMSCS) program. The original lectures are freely available online and presented by Dr. Spencer Rugaber, as open-source content.

  There are over 35 quizzes that comprise 5% of the grade based on the lectures, designated as a form of “padding” the grade according to the TAs. There is a group project with group members organically organized through forum posts, usually within the same time zone to facilitate availability for online meeting times. These groups and their projects tend to have high variability in quality and effect, such as determining leadership, which skills each team member possesses, how they can interact and allocate time over the semester while most (80%) of the students are working in industry, etc. There are also two open-ended exams, with the second exam optional depending on the student’s performance on the first exam. Given these course requirements, and the academic rigor of a “superelite” Institute of Technology like Georgia Tech, the barriers to learn this course’s material in a formal, academic setting are relatively high. However, the course material, especially the lectures, are valuable enough to others that a condensed, edited, visual reference and guide would be a godsend. Effective and efficient software systems form an area of critical need in our society today, and software architecture training provides a tried-and-true method of enabling the development and execution of these large-scale software systems. These systems can span millions of lines of code, tens or hundreds of modules, and model and run large, essential infrastructure projects like air traffic control and routing or power plant operations.

  In industry, software system architect is generally understood as the role superior to that of senior software developer, regardless of title. The software system architect works with management and customers to define technical specifications and high-level software solutions, while also directing a team’s or teams’ development throughout the project from a technical standpoint. I took this class to get a head start on the next level of work that would be available to me after working as a senior developer for some time; it is difficult to get training or experience in industry on software architecture topics unless the company promotes or authorizes an individual, while Georgia Tech provides this class to anyone enrolled in the OMSCS program – it is never full or impacted, with no prerequisites beyond admission. These notes could also provide awareness and perspective to those working at technology firms in the industry who do not work with code directly, to give insight into how these applications are made from a high-level perspective.

  For more information, check out my Summer 2026 update video! https://www.youtube.com/watch?v=5ZJ4rllLG7o

  Rating: 4 / 5Difficulty: 3 / 5Workload: 16 hours / week

• l6Z2LEWR4pUHmqtA4hR5Kg==September 3, 2026spring 2026

  Advanced Operating Systems

  This review and course summary was taken from the foreword of my AOS course study packet, available on Apple Books (https://books.apple.com/us/book/advanced-operating-systems/id6799153989) and Amazon (https://www.amazon.com/gp/product/B0HDK5CCCJ?storeType=ebooks&sr=8-1):

  This study guide consists of my notes for the upper-level Georgia Tech graduate school course entitled “Advanced Operating Systems” (AOS). The motivation for compiling, editing, and publishing these notes was primarily to pass the course; a little more than half the class (over 300 students enroll per semester!) currently receives an A or B each semester, with a B as the minimum grade required if in the Computing Systems specialization, which I am. The notes were synergistic with my coursework; I was able to get an A thanks to these notes, along with dedicated efforts on projects. The proof is in the pudding; I started these notes a month and a half early in the summer before I started the class, due to the large span of material covered in the course, and was able to get an average, passing grade on the first exam (of three) by only reviewing these notes for the topics covered over the course of four days before the exam. The latter two exams took about 10 days to prepare for and get A's on, with other activities going on. The exam format should be discussed in some detail to add context to the utility and purpose of these notes – there are three open-notes, open-internet, open-forum discussion (yes, you can ask and post questions to other students about the exam questions during the three-day exam period) spaced evenly throughout the AOS course. Each exam has questions posted before a weekend, then the students have two days to prepare their answers and a third day to “fill out the form” in the space of an approximately two-hour, proctored “exam” session. The final exam for this course, with an average amount of questions and points, required 5 pages of notes and writing to address all the questions adequately, amounting to a keyboard and memory performance during the exam session. These open-ended, short answer exams are notorious and widely considered the worst part of the class, requiring looking up trivia and facts and logical reasoning based on 48 papers throughout the span of the course. So this relatively concise note packet should serve as a supplement to the lectures and papers, enhancing understanding and making the studying, review, and lookup process that much easier. Recover hope, all ye who enter here – the notes packet and the course are finite, regardless of how they seem on the outset. Good luck! I’ll be waiting for you on the other side!

  For more information, check out my Summer 2026 update video! https://www.youtube.com/watch?v=5ZJ4rllLG7o

  Rating: 4 / 5Difficulty: 4 / 5Workload: 28 hours / week

• 0qNdCyeDuSTnuVthrbLt9g==September 1, 2026summer 2026

  Introduction to Graduate Algorithms

  Anyone who left a positive review is either the TA or the professor himself. The videos are outdated; it is another professor (Eric Vigoda) and not Professor Brito himself. They reuse the same video for a decade now. I expect I get more than how much I paid for this class. The TAs are not qualified to teach this class and I am 100% sure they will not be able to pass basic algorithm class themselves based on their background. They are racist as well.

  Rating: 1 / 5Difficulty: 5 / 5Workload: 20 hours / week

• kPRbSxTFjIkenr0EYxqjcQ==August 28, 2026summer 2026

  Introduction to Cognitive Science

  The course primarily involves a large term project which has an expected workload of >70 hours. It is quite daunting and open ended, requiring you to pick a topic, find your own research papers, and synthesize findings. Grading is pretty lax though.

  The lectures in the course are easy to follow, but the reading assignments are endless and impractical. The quizzes are surprisingly difficult, but they are open-everything including AI.

  Overall, unless you love reading and research, you may not enjoy this course. In terms of workload, it's extremely front loaded with the second half of the course is very light. I was able to get ahead in my project and have some time free at the end.

  Rating: 3 / 5Difficulty: 4 / 5Workload: 8 hours / week

• WpUxDolCMCvJ2Yx1SPJJMg==August 28, 2026fall 2025

  Database Systems Concepts and Design

  I’m almost halfway through the OMSCS program, and this course is still my only C in the program so far. I took it as my first OMSCS course, before I knew about OMSCentral or had much information about which courses to choose.

  The most applicable things I got out of the course were learning some SQL and getting experience working on a group project that included front-end development. As someone interested in full-stack development, I do think a database course is valuable and that understanding SQL and relational databases is important.

  However, I had several issues with this particular course. The material focuses almost entirely on relational databases, so the overall scope felt limited compared with what I would expect from a modern database course. The exams were also one of my biggest frustrations. Many questions felt intentionally tricky, where understanding exactly what the question was asking could be harder than demonstrating your understanding of the database concepts themselves.

  Another issue was the exam integrity. After one of the exams, I came across a diagram from what appeared to be a previously used exam circulating online.

  Overall, I think taking a database course is especially useful for students pursuing full-stack or backend development, but I have a hard time recommending this particular version of the course. If you’re considering it, I would personally wait for a course redesign or major refresh.

  Rating: 1 / 5Difficulty: 3 / 5Workload: 14 hours / week

• Z9cH0A80pMRn1edqXMOwdQ==August 27, 2026summer 2026

  Special Topics: Intro to Research

  I took this class in Summer 2026. The class consists of two papers that you work on simultaneously: an individual project proposal and a group literature review. You make updates to those papers every week and peer review other groups/individuals in between. There are also module quizzes but these are just 5 questions quizzes with unlimited attempts going over the lectures so you can get full points there.

  This is a relaxed class that you may enjoy if you want to research a specific topic in more detail, take a break from programming heavy classes, improve/maintain you GPA, or improve your academic writing. The course is graded very leniently. It will be difficult to get anything less than an A unless you do not turn in the assignments.

  I usually dread large papers and procrastinate but the weekly logs help you say on track and makes each portion manageable. The peer reviews can get boring and monotonous over the semester but not a big deal.

  If your goal is to perform real research immediately and or get a publication consider taking a shortcut and jumping straight into a research group or doing an independent study (8903 course). This class is not strictly necessary but can help learn academic writing or create a proposal to present to a professor for faculty sponsorship.

  Rating: 3 / 5Difficulty: 1 / 5Workload: 4 hours / week

• bDapZAFAaavKq2nZkErCDQ==August 25, 2026summer 2026

  Database System Implementation

  This course offers a strong, hands-on foundation in the low-level mechanics of database systems, using C++ to implement core components such as indexing structures, schema management, tuple storage, SQL query processing, partitioning, and synchronization mechanisms. As a software engineer, it's valuable to understand these internals.

  However, the course doesn't cover the practical, applied side of using databases — for example, how to set up and deploy a database, configure it for different real-world use cases, or analyze and tune SQL query performance. Only some basic lectures touch on these topics.

  This is a heavy coding course, and solid C++ skills are essential. The assignments are not easy — they can take a significant amount of time to figure out and get working correctly.

  The practice exams, midterm, and final are all closed-book. I found them manageable but not easy, since the content draws from research papers, lectures, and the textbook alike.

  Overall, this is a good course for understanding how databases work at a low level, but it requires substantial coding effort. I was hoping for more exposure to practical, real-world database usage, which this course doesn't really provide.

  Rating: 4 / 5Difficulty: 4 / 5Workload: 16 hours / week

• HBSwSIlY4W/ObBOpTutzyg==August 24, 2026summer 2026

  Foundations of Computer Graphics

  Overall, this was a great course. The course reaches a good balance between theoretical background and practical application. The topics are really interesting, especially the ray tracing module (which is half of the course), you get to learn a lot of the specifics for implementing the algorithms, then apply them in a friendly way in the assignments and finally you will see it reflected in your screen. The grade is divided into 5 assignments (80%) and many quizzes about the lectures (20%).

  Assignments: They provide you with a good starting code, so that you can focus on implementing the topics you learned in class. They also provide good public and private test cases to test individual functions, which makes it easier to know you are on the right track. Finally, you are also evaluated based on images of the target scenes you need to generate. The assignments are programmed in Java using the Processing library.

  Quizzes: I find them easy and a good refresher of the lectures

  Some minor improvements:

  • Processing can be good for learning, but it would be nice to learn a graphics API that is used in the real world. At least for some modules.

  • I got less points in one question of the last assignment because I didn’t implement the function in the exact same way as the lecture described it, even though mine was clearly that it was mathematically equivalent to it (and the final output was correct). It didn’t end up affecting my grade but, if you are taking this course, be careful with that.

  Rating: 5 / 5Difficulty: 3 / 5Workload: 10 hours / week

• 17lm/XFuCbdcqY63XAcCMQ==August 24, 2026summer 2026

  Data Mining and Statistical Learning

  Background: I have a BS in Math & Statistics and took this class Summer 2026 to fulfill the Analytics MS stats elective and review some models.

  Effort: I didn't watch any of the lectures, referenced the lecture notes for some quizzes, and completed the homework and project by referencing online resources (geeksforgeeks, statology, etc.). I estimate spending 10 hours/week in summer on this class, the main effort going toward making my homework reports accurate, professional, and easy to read (LaTex).

  Pros: You can go as in depth as you'd like with any concept. You get to choose your project and work alone. The peer reviews are great to see how my work was interpreted by others, what worked well to communicate concepts, and what could be improved.

  Cons: I'm concerned by half of the reports I read. The other half were great and I could tell the students were genuinely giving a full effort into the assignment and reviewing others. However, some were like a draft I would read for a middle school paper. My college homework had much higher grading standards and effort; the grade distribution for the assignments still was centered around an A. As well, the quizzes didn't require Honorlock and mirrored the knowledge checks. I feel there isn't a baseline for understanding concepts -- mid effort = A. Specifically, if the quantitative concepts are covered in lecture, then I think the homework and project should require students to demonstrate their mathematical understanding.

  Recommendation: Take this class if you want an easy A, need to fulfill degree requirements, or want an open-ended project.

  Rating: 2 / 5Difficulty: 1 / 5Workload: 8 hours / week

• Qd5E2FDh5FoEBiot2Zil0Q==August 24, 2026summer 2026

  AI, Ethics, and Society

  CS 6603 is one of the lighter courses in the program, and it's a reasonable choice if you want to balance a heavier course or recover some breathing room in a summer term. No exams, no major coding project, the work is assignments across four modules covering data and society, statistics and the pitfalls of big data, fairness in AI/ML, and a final applied piece on quantifying and mitigating bias. Easy to get Grade A if studied very well. Worth taking if you want a manageable term or genuinely care about the subject. If you're looking for deep technical ML content, this isn't it.

  Rating: 5 / 5Difficulty: 2 / 5Workload: 1 hours / week

• QXVJyKsMsSwQ8RZ4khjCCw==August 22, 2026summer 2026

  Software Development Process

  Coming from a background where I never did CS undergraduate classes, I was a mechanical engineering BS graduate with 3 years of experience as a data engineer, I found this course to be actually very useful. Again, SW engineers may disagree but I think this course is very valuable for people who did not do CS undergraduate degree.

  This course goes over github, automated coding testing, javascript and android application creation. All of these I found to be extremely applicable to my industry and really helped me brush up on my github skills.

  Overall, the class wasn't hard. The hardest week was where assignment 6 and the the team project was due at once, which brought me up to like 15-20 hours of work in that week, most other weeks it takes like 8 hours usually. Except for the first week of the individual project, I spent up to 20 hours for that one week but then less than 5 for the last two weeks.

  All in all I thought the class was useful, applicable, and not that bad if you had a decent group. Only criticism I would have is the team project and the peer reviews where 10% of your grade is dependent on how your teammates rate you out of a 100. So if you get a crap group...good luck. I ended up with a decent one, of course you will probably have last minute procrastinators forcing you to stay up to 3am before the team project is due. But overall everyone in my team contributed even if I had to stay up to 3am before the team project was due. It wasn't that hard anyway, someone who knows SQLlite and javascript could do it all on their own faster than having to coordinate with the team. Because I was new to both, I learned a lot, and it took longer like ~15 hours per week during the 2nd deliverable, but experienced engineers could finish the whole thing in less than 20 hours over the 3 weeks.

  Overall, I'd recommend this class over GA any day of the week.

  Rating: 4 / 5Difficulty: 4 / 5Workload: 12 hours / week

• w+ZN2+tIOB2J+M4Dlz0b/w==August 22, 2026summer 2026

  Educational Technology: Conceptual Foundations

  Superb course. I think it is more about conducting an academic research and writing a paper than developing a technology for education. It is really self-paced and self-directed, as in you get to choose what you want to dive into, and work on a project of your choosing.

  Rating: 5 / 5Difficulty: 3 / 5Workload: 12 hours / week

• w+ZN2+tIOB2J+M4Dlz0b/w==August 22, 2026summer 2026

  Human-Computer Interaction

  Pros:

  • It gave me very useful knowledge and skills on creating and evaluating designs.
  • Lectures are numerous and fun.

  Cons:

  • A bit too much workload -- 3 homeworks, 4 quizzes, 2 tests, 1 individual project, and 1 team project, all in one semester.
  • Little time to digest the material -- I did so many assignments, but meaningful evaluation is lacking, and I felt like I was just trying to complete the next assignment, rather than letting the concepts sink in, ask questions, etc.
  • Honorlock.

  Pitfalls I fell into:

  • Letting Honorlock annoy me too much -- When taking quizzes and tests, I thought how terrible the Honorlock experience was, and I failed to concentrate 100% on the tests.
  • Forgetting the overleaf link -- In my team project, overleaf link was missing from the submission, and we got a 10% deduction on the assignment.

  Rating: 4 / 5Difficulty: 3 / 5Workload: 15 hours / week

• N7TeG3r8JXATtUnI8xlOZg==August 20, 2026summer 2026

  Game Artificial Intelligence

  This was a great course and introductory to AI concepts. This may be more manageable for a full semester but this is not an easy class that you can just walk through in 12 weeks.

  The class environment was the best I've experienced so far. Lecture's are long but informative if you're interested in game development. Do the quizzes right after you watch them because they're fresh in your memory. They give you until the end of the semester but if you wait until the end you end up realizing there's a lot of content to remember.

  The TA's encourage use of discord and student's are allowed to help each other (within reason). Highly recommend this class even if you don't like game development.

  Rating: 5 / 5Difficulty: 4 / 5Workload: 12 hours / week

• KhBR/67HXmW3JAR+at3Pvg==August 20, 2026spring 2026

  Machine Learning

  Final Result: A

  After giving up on this course in the summer of 2025 (big mistake do not do this in Summer), I dared to take this course again in the Spring of 2026 and I can attribute my success in this course two things directly. Google’s Notebook LM and some saint who uploaded all of the lectures to YouTube making it possible for me to dump these Youtube videos into NotebookLM and generate drive time podcasts for the same that sort of abstracted away all the Math from the lectures and gave me a distilled anecdotal learning experience from the video lectures. It also helped me that the course allowed the use of AI generated code so the majority of the time I could focus on learning the course content and following the instructions in the assignment prompts to nail the report down. The course faculty also give you back half of the points for the first 3 reports if you resubmit with all the improvements they have pointed out and this was another main reason why I made it with an A.

  This was one of the most satisfying courses you could think of. The course content was foundational and dived deep into the four main foundations of Machine Learning methods, SL, Neural Networks and Optimizations, UL and RL. Some part of the course could also be shortened by removing away the RL pieces given there is a entire separate course for this giving us a little bit more time to dive deep into the course specifics.

  The exams were also kinder as you had the opportunity take cheat sheets and could repeat the quizzes so the focus of this course was always on the learning aspect and never on worrying about the students using AI for cheating or rote memorization.

  Rating: 5 / 5Difficulty: 5 / 5Workload: 20 hours / week

• KhBR/67HXmW3JAR+at3Pvg==August 20, 2026spring 2026

  Introduction to Cognitive Science

  Final Result: A

  I finished the course KBAI in Fall’2025 and I wanted to take up this course to both dive deep into the foundational knowledge of cognitive science and also make sure that I have slightly lesser intense course with my main course for this semester which was Machine Learning. This course was a very writing focused course and we did not have any exams or quizzes for the course. To pair this with Machine Learning was actually a double edged sword because this course had a writing deliverable every week or every other week and I had to scramble to finish these reports while the larger reports from ML were due.

  A good time to take this course would be before a writing heavy course so that you get used to the writing and research aspects of a rigorous program like OMSCS. My favorite part of this course was definitely the rigor of the course material. Special shoutout to Professor Ashok Goel and Professor Joyner for making the course content engaging and digestible. You do not need to go back and watch the videos over and over while you try and decipher the course content. Listening to the lectures two or three times is more than enough to internalize all the important topics that are mentioned in the course.

  My favorite part of the course has to be the final project. This was almost like a masters thesis where we defended a hypothesis we had by distilling other research for a lack of better way to explain it. The quality of research and hypothesis support was definitely masters level and I could totally see myself publishing something of this quality some day imho.

  Rating: 4 / 5Difficulty: 1 / 5Workload: 10 hours / week

• KhBR/67HXmW3JAR+at3Pvg==August 20, 2026summer 2026

  Natural Language Processing

  Final Result: B (87.9%) Semester: Summer 2026

  In this day and age, with Large Language Models being the core drivers of enthusiasm in everything from financial markets to research corridors, this course helps you dive deep into those neural networks giving you an eagles eye view of the wonder that transformers are. For every module that is being covered in the course, we begin with very simple foundations in probabilistic methods and then graduate from there into diving deep into advanced neural architectures that help us achieve the same goal.

  The meat of everyday LLMs is laid out in the fourth and sixth modules which cover everything language modeling and generation from encoders and decoders to attention mechanisms. These two modules should be a full course in themselves given the dense amount of knowledge they contain. Every single lecture that is a part of this module unravels one layer of complexity from LLMs. I think at some point in time before I graduate I would like to build all these different neural architectures that are so foundational so everything we see today in AI applications.

  The second half of the course has the infamous Meta AI lectures and these really are bad and only then do you appreciate how good Professor Mark Riedl is. In a sharp contrast to the professors lectures, these Meta AI lectures are uninspiring and information dense with absolutely no intellectual simulation. The content delivery is super monotonous and if not for the notes from here. I would not have understood anything from the course.

  One of the weirdest things I felt about this course was that the quizzes are a single attempt and nearly 40% of the grade comes from quizzes (10%) and exams (30%) so the overall burden on memorizing certain facts from the course becomes high which takes away the fun from such interesting topics. If I could re-design the course, I would definitely make it more like CS7641 and have some research angle driven into it. The final project is indeed one of the hardest I have seen and because you cannot use AI to get through it you are left with a lot of diving deep on the internet to find answers to questions you may have about the topic you are implementing from. Maybe having another project of this level in the middle of the course would be way more fun than having single attempt exams where some questions are definitely out there to trick you (with the use of weasel words or such).

  I also did not feel there was a lot of social camaraderie from the rest of the class in this course. I made some friends in my earlier courses but given the strict deadlines with a quiz due every week and reams of videos to go through and no chat-forum like interface for students to chat about most of the interactions were limited to asking questions about the course and answering them alone. I wonder if this is a summer course tbh, I think having four more weeks will enhance the learning experience and make it easier for students to digest the course material. I am writing this review 3 weeks after finishing the course and I am sorta worried that I may have forgotten most of what I learnt.

  Rating: 4 / 5Difficulty: 3 / 5Workload: 12 hours / week

• B4Rg+t5El3svbUx4Dnp1Nw==August 20, 2026summer 2026

  Introduction to Graduate Algorithms

  The course itself isn't that hard and is pretty interesting, but there's certainly frustration with the way exams are graded and the things that are nitpicked. A simple mistake can lead to an 8-12 point deduction on an FRQ. Unless you can't take the grade hit, I would stay put even with a bad exam 1 grade. The class becomes increasingly more mechanical with every exam and worst-case scenario you know the content better for a second go-around.

  Thoughts/Tips:

  • Watch the lectures ahead of time if you can, especially if you are taking the class over the Summer.

  • The class uses side-camera proctoring, so be prepared to buy a webcam and stress out when there's a network hiccup during the exam. Honorlock is and probably always will be awful.

  • Go through all the walkthroughs of problems (both PDFs and videos). That includes the content/format quizzes, even if you did well on them. Some good info in those.

  • Attend/watch the recap of the office hours. Prioritize TA ones over professor. Brito is a nice guy, but he sometimes takes shortcuts for solving problems which won't serve you well on the exam. Plus, the TA ones cover common pitfalls + good guidance tips for each exam.

  • Have the format for exam questions memorized. This is covered on the pinned posts in Ed Discussion.

  • Don't rely on AI too much. You cannot modify the algorithms themselves for exams 2 and 3 and the AI will try to do so unless you tell it that it cannot.

  • Do the Textbook readings. Not every practice problem in the book is made equal. Focus on those that can be formatted the way that is asked in class. I personally thought the "challenge" ones weren't worth doing.

  • You are setting yourself up for failure if you don't do the written HWs. I would do these multiple times. They are the best prep for the exams. Unfortunately, you don't know what you're doing wrong until after they grade the HW. Ed Discussion + office hours are good resources to see how others approach the problem.

  • The programming HWs are not particularly useful, unless you can turn them into written problems. This isn't a coding class.

  • There's a new optional quiz 8 that allows you to get points that you may have lost on past content quizzes, but try to avoid doing it if possible. For some reason, it's proctored unlike the others which felt stupid and unfair and that's coming from someone who didn't do it.

  • For the exams, write down important aspects of the problem. What are the inputs/outputs? What is the goal? Jotting down thoughts is useful.

  • Separate sections into different blocks and clearly label them. Too many students "hope" that the grader will see all aspects of an answer.

  • Write down the MCQs + choices during the exam. Some are too long to write word-for-word but you don't get to see what questions you missed after the exam is graded and the TA feedback is way too general and unhelpful.

  • Exam 1 was definitely the toughest (which seems intentional) - especially the dynamic programming problems. All that you can really do to prepare for DP is practice. For the divide and conquer, make sure you are familiar with binary search and all aspects of the algo. It will probably come in handy.

  • Exam 2 had a lot of content. Create flashcards and lump algos together based on common inputs/outputs/runtimes. Come ready with some toy graph examples to test solutions. Consider whether you need to check a graph in both directions/account for multiple sources/sinks.

  • The NP unit is super formulaic. The input transformations are generally similar based on problem type (like adding clauses/variables for satisfiability, gadgets for graph problems). For graphs, build an induced subgraph for both verifying solutions and constructing the output - that way you aren't removing things that were part of the original graph. Also, don't forget the NO condition and make sure your correctness proof is arguing converse implications. Remembering like five of the algorithms from the list they provide is probably sufficient.

  Good luck!

  Rating: 4 / 5Difficulty: 4 / 5Workload: 20 hours / week

• dTmnol/8G3Ie0m+X+uai5Q==August 20, 2026summer 2026

  Machine Learning for Trading

  People planning to take the course needs to pay attention to the age of the reviews. The course has gone through some serious overhaul last few years, and the experience may be very different than when the class was first put together.

  Coming into this class, I had no prior knowledge in machine learning and finance. I've found the lectures and the assignments interesting as they provide introductory ideas in finance and machine learning using real stock data from the past. I've seen few people complaining about how introductory the lessons are, but I think that's an unfair criticism, because again, the class is designed to give an introduction on ML using real world examples for those who hadn't had an experience in the field.

  The problem is that the course felt very 'patched up'. Some of the readings weren't reinforced by projects or assignments, but they were expected to be thoroughly memorized for the exams. Therefore, it lacked depth and basically became details that I memorize for tests and forget right after. This added a lot of extra time and stress, as the tests were closed-everything. I think many of the readings were added later on so that the class is somewhat up-to-date with modern technology, but they lack proper lesson plans to be a meaningful learning experience.

  Most students do well on projects so much of the grading scheme felt like it was harshly penalizing to reduce number of students that get A. For example, having the plot in different colors or not generating the file in a correct format can reduce your points by 20%. However, there were a lot of strikethroughs, typos, and ambiguities, so you could easily miss some simple instructions if you weren't being extra careful.

  Workload was also very inconsistent. All assignments were given one week to complete. While some projects, (project 3 and project 8) took way longer, some projects were done in just a few hours. This may be because I took the course during the summer, so the pacing could be different for other terms.

  Overall, I felt that the course was not well-organized. I also thought that it was more difficult and time consuming than I expected. My personal opinion is that instead of forcefully adding new contents, it should focus on enhancing and updating the original lessons. As quizzes don't really add any value at this point, they could be utilized to introduce new ideas and technologies at a higher level.

  Rating: 2 / 5Difficulty: 4 / 5Workload: 20 hours / week

• r8FIhsWYwRZQ0iw4gmpt1A==August 19, 2026summer 2026

  Machine Learning for Trading

  DISCLOSURE: This was generated using a LLM after providing the following prompt and providing responses to the requested interview:

  I'm looking to write a comprehensive review of my experience in CS7646 and would like your assistance. I think I'd like to break it up into a review of the lectures (specifically the three distict phases), the readings, the projects, the quizzes, the exams, and the TAs. Can you act as an interviewer to capture my thoughts on each of these and pull them together after I finish?

  I have proofread this to confirm that there are no material misrepresentations of my interview responses. My final score in the course was an A. My background is a BS in Computer Science and significant professional experience as a Software Developer/Engineer in non-AI/ML projects.

  ---

  I finished CS7646 feeling largely unfulfilled. There were parts I enjoyed (a specific lecture/project combination) but I would not recommend the course in its current form without a substantial overhaul.

  The course is broadly divided into three phases: learning Python/NumPy/Pandas/Matplotlib, learning basic finance, and learning introductory machine-learning techniques. The finance material felt appropriate for CS students who may have little market experience. The programming portion, however, felt too elementary for a graduate CS course and occupied more time than warranted. Much of the lecture material is also approaching a decade old, and that age now shows in the examples and software environment.

  The ML lectures were more interesting, but generally emphasized what algorithms do and how to implement them rather than exploring why they work in much depth. That may be reasonable because OMSCS offers a dedicated ML course, but it makes the readings feel particularly odd: several external texts go far deeper into ML theory than the lectures ever do, while the course-associated text largely repeats the lectures. The deeper readings seemed useful mainly for exams or for students who independently wanted a broader ML education, rather than because they were well integrated into an applied ML-for-trading course. There were also lectures that felt disconnected from the course's stated focus. For example, stock options were introduced without meaningful follow-through into technical analysis or ML.

  The projects varied considerably. Several were useful preparation or finance-domain exercises but contained little substantive ML or graduate-level CS. One project effectively demonstrated weaknesses of a particular ML technique, but did so by expecting students to discover important lessons experimentally rather than teaching them first and using the assignment to reinforce them. There was one standout project that had genuine implementation nuance. The associated lectures provided a useful conceptual model, and the project reinforced the lecture material directly. It was easily my favorite part of the course and the best example of what I wish the rest of CS7646 had been.

  The capstone was a disappointment. It integrated earlier work, but did not feel like a major technical challenge. Students were also largely constrained by decisions made in earlier projects, limiting the opportunity to use the project to discover better alternatives. That connects to my biggest criticism: too much of the meaningful learning happens through independent research and experimentation outside the actual instruction. Graduate students should absolutely learn independently, but when that becomes the primary mechanism for learning, it raises the question of what the course itself is adding beyond structure, grading, and degree credit.

  The quizzes added little. They were short, open-resource, low-stakes, and easy to score well on with basic care. The exams were stranger: they were not recall tests, but often involved enough indirection and synthesis that it was difficult to tell exactly what knowledge or skill was being assessed. The course did, however, do an excellent job with exam feedback. Statistical validity information and individualized conceptual feedback were unusually thorough, even if difficult to parse without seeing the original questions.

  My experience with the TA structure was also poor. The TAs were responsive and available, but I found answers about rules, rubrics, and assignment requirements frequently vague rather than clarifying. Course staff explicitly defended some ambiguity as realistic because real-world requirements are often unclear. That is true, but real-world work also usually provides opportunities to clarify, iterate, and correct misunderstandings. A grading environment that intentionally preserves ambiguity without equivalent opportunities to recover from it does not reproduce that reality particularly well. The TA-produced project overview videos also added little beyond restating published instructions. Aside from grading, I personally received essentially no educational value from the TA structure.

  My strongest ethical concern was the use of required mid-course surveys evaluating the course. Students were penalized for non-participation while the people being evaluated still exercised grading authority, and there was no clear information establishing whether responses were anonymous, confidential, or withheld from staff until after grades were final. I am not alleging retaliation occurred; the problem is that the structure unnecessarily creates the opportunity and appearance of a conflict of interest. This could easily be avoided by collecting evaluations after final grades, using an independent third party, making participation optional, or offering minor bonus credit rather than penalizing non-participation.

  CS7646 contains pieces of a much better course. The finance introduction is useful, the ML material can be interesting, the exam feedback is strong, and there are specific units that show how effective the course can be when lecture and assignment reinforce each other.

  Unfortunately, that felt like the exception rather than the rule. Without significant modernization and pedagogical restructuring, I would not recommend it. Had it not also provided three credits toward my degree, I would have difficulty concluding that the educational experience alone justified the time and tuition.

  Rating: 2 / 5Difficulty: 2 / 5Workload: 12 hours / week

• Pqc0oBP00i24fWaUAIuZDQ==August 19, 2026summer 2026

  Software Development Process

  Full disclosure this is my 10th class in the program and I only took it to avoid the Graduate Algorithms doom-loop of failure. I figured I can try GA once more and if it doesn't work out, I can drop GA and just apply for graduation in a different specialization. This seemed like an easy option or plan B since I have all the required courses either way.

  I have decades of software development experience and didn't expect to learn much from this class. I wanted an easy summer class and this fit the bill. But I actually did learn a little bit, especially the testing modules and even used the partition testing material at work and it was quite effective.

  The other part of this class that was surprisingly worthwhile were the instructor's AMA. I forget his name, but his answer to questions in the Ed Discussions on the AMA threads were amazingly good. There were a gazillion questions related to AI and its impact on the field of software development and I found his responses to be amongst the best I have ever read.

  The office hours were essentially non-existent. I joined one and there was like one other person there and you had to pre-submit your questions ahead of time. Also, the TAs were not very responsive and all of the grading pretty much happened the final week. Which was frustrating because you had no idea where you stood all semester until the final weekend before grades were due.

  Watch out for assignment #6. It is tricky, worth a lot of points and Gradescope only gives you a provisional grade. A relatively simple assignment, at least on the surface, ended up being deceptively more complex and graded more severely than I anticipated. I ended up with a 50% on it and it brought my grade down from an A to a B - which for my purposes was just fine.

  The group project goes over Android and it is actually useful if you want to learn Android development. I had a really good group and we did well on it.

  Some of the assignments have a lot of strange requirements in them and they are like 10 pages long of how to do something. I goofed a few of them up just getting confused what they were asking for. I think the deliverables and instructions for these assignments should be simplified just a little.

  In the summer the class is compressed a little so there is no downtime. No exams which is nice. The class was kind of a sleepy class in the summer. Not a lot going on in ed discussions which I appreciated. Just do the assignments as they are due and you really do not need to check in very often. The course just kind of runs itself.

  Overall I spent minimal time in the course, got a good enough grade, didn't learn much given my background, but did find a few things useful and picked up a couple new skills. All in all, I am glad I did it and it isn't the most well-run course I have taken in the program, but for some people, I think it can fit the bill.

  Rating: 3 / 5Difficulty: 2 / 5Workload: 6 hours / week

• UmpoA1qvCf95whP2KKtuNQ==August 19, 2026summer 2026

  GPU Hardware and Software

  Background: I took GPU HW & SW as my third course, having taken GIOS and HPCA previously. I have always been interested in how GPUs work, and felt that being able to program in CUDA is a great skill - hence I decided to take this course. I ended the course with an A (98%).

  Pros:

  • Project 2 was one of the most fun/interesting projects I've done so far. (I've taken GIOS and HPCA).
  • There are weekly Paper Reading & Discussions, where you can volunteer to present a paper on a topic to other students and TAs. I thought this was a great idea, and more courses that are paper-heavy should implement something like this.
  • I would say this was the strongest set of TAs I've encountered so far. All super responsive, friendly, no ego, and very helpful in their replies to technical questions.
  • I noticed that the other students I did the course with were super enthusiastic and passionate about the subject. Not sure if this is GPU specific?

  Cons:

  • Dr Hyesoon Kim is definitely a leading expert in this field - I've watched a lot of her talks and interviews. However, her lecture delivery and teaching style is quite dry (reading off slides), and borderline hard to follow.
  • The other projects were meh. Proj1 is mostly an intro to using PACE. Proj3 doesn't have much learning value imo. Proj4 was a bit more interesting, we got to learn and implement Attention. Proj5 was basically a Leetcode quesiton.
  • A lot of the questions in the Final Exam are adaptations of the quiz questions.

  Rating: 4 / 5Difficulty: 3 / 5Workload: 16 hours / week

• M7d044xNxI54WMxNYBXbmg==August 18, 2026summer 2026

  Natural Language Processing

  GOOD PART

  First half of the course (the part taught by the professor) is really really good. He teaches attention mechanism etc from a perspective which makes it feel so intuititive. It is totally worth taking the course just for this part.

  BAD PART: The other half (taught by Meta guys) is horrible. They might as well would have just given us a written summary for that part.

  Also the course management is horrible. They are always so paranoid about students cheating (like they don't even provide you slides or lectures to download) that it feels as if they are going through some PTSD phase.

  Rating: 4 / 5Difficulty: 3 / 5Workload: 4 hours / week

• M7d044xNxI54WMxNYBXbmg==August 18, 2026spring 2026

  Computer Graphics in AI Era

  Really great course. I loved it. Learned so many new things. Professor and TAs are also super nice.

  Rating: 5 / 5Difficulty: 3 / 5Workload: 15 hours / week

• 4v+GPNibbZHV4cDoKVlvvg==August 15, 2026summer 2026

  Digital Marketing

  It's really easy, no doubt about it. A lot of the content is common sense; it's just common sense that a lot of folks forget to apply, and so is helpful to hear out loud. So I do think the content is worthwhile, even though it seems obvious at a glance. I do wonder whether the prevalence of AI generated content will make some of this theory outdated in a few years, though.

  The major case studies are fairly easy to write casually in an afternoon (though one required some mathematical calculations, and that one was an outlier), and the mini case studies can be done in like... 20 minutes. The lectures are <1 hour per week, and usually <30 minutes. The reading is what will take the longest, and some questions do in fact come from the book for the exams, so I think if you're actually wanting to learn the content you should do the reading.

  All in all, easy, but there is valuable content.

  Rating: 4 / 5Difficulty: 1 / 5Workload: 5 hours / week

• 4v+GPNibbZHV4cDoKVlvvg==August 15, 2026summer 2026

  Computer Animation

  This feels like more of a math/physics class, or SIMULATION class, more than a computer science or ANIMATION class. Pretty much every project is "implement this mathematical algorithm that we told you about in lectures". All animation libraries are handled for you, so it ultimately just comes down to a lot of interpolation formulas, physics, etc. That's not to say there's nothing interesting here: Keyframes, collision detection, etc are discussed a bit - but Video Game Design, which I'd already taken, talked about a lot of these same concepts already, leaving mostly mathematical implementation.

  The TA's for the course were fantastic, and clearly very engaged and happy to be helping out students. So I have a lot of respect for how they're running the course. Assignments have a lot of extra bonus points available if you're willing to apply a deeper level of understanding, and the vast majority of the grade (85%) is covered by assignments and quizzes that have unlimited attempts. And even the exams were really easy, so long as you studied even a little.

  I'll note that if you never took linear algebra or differential equations, you'll have a little bit of catching up to do; but it's all stuff you'd learn within the first couple weeks of those math courses, so nothing too bad.

  So all in all, I think the course is run quite well. It's just that you should consider whether implementing mathematical interpolation/simulation algorithms in numpy is interesting to you. If so, you'll love this class, but it wasn't for me.

  Rating: 3 / 5Difficulty: 2 / 5Workload: 9 hours / week

• AJ0BdNtaNSOhCwCAM25qXQ==August 14, 2026summer 2026

  Information Policy and Management

  Most of the reviews here are pretty dead on. Majority of your time in the course will be spend reading/watching lectures and memorizing a ton of relatively useless information for the free response (closed book, proctored) quizzes. The quizzes aren't that difficult and the TAs seems to give a bit of grace in your answers, but you really do need to memorize the slides. There's hardly ANY talk of actual InfoSec concepts, and the small amount you do cover is fundamental at best. The worst part of the class was easily the final project which is a 50% of your grade - 20 page paper and PPT presentation, random group assignments. I got 2 complete dud teammates so me and another student essentially did the entire thing ourselves. Professor Rogers is kind and responsive, at least. All in all not a tough course, just really depends on how responsible of a group you're assigned.

  Rating: 3 / 5Difficulty: 2 / 5Workload: 5 hours / week

• BrHCAwk8rHVMLMavnc4MiA==August 14, 2026fall 2025

  Computing for Data Analysis: Methods and Tools

  Overall, really enjoyed this course. This was the 3rd class I took in the program as someone that came in with very little coding background. I felt like it was very challenging but also quite rewarding. The professor and team of TAs are great, the lecture is interesting, and the homework teaches you a lot.

  If you are coming into this course with a solid python background I think that it shouldnt be too hard (probably <10 hours a week of work). However, if you come into it without much experience in python you might have to put in a bit more time.

  Best advice is that the homework will get easier as you go and to fully utilize the practice tests available. I found that I had to complete all available practice exams before each test. I would do a few without any time limit and then the last bunch in a timed setting similar to the real exam.

  Rating: 5 / 5Difficulty: 4 / 5Workload: 17 hours / week

• BrHCAwk8rHVMLMavnc4MiA==August 14, 2026spring 2026

  Data and Visual Analytics

  Overall a pretty interesting course but quite challenging. I have come into the OMSA program without much of a coding background and definitely struggled with the coding in this class. The homework assignments are pretty front loaded as they are reasonably complicated D3 problems up front. As someone with absolutely no exposure to D3 beforehand, it took a lot of initial work. Content also does feel quite broad. A lot of information is covered and any further learning will be up to the student. I found myself a bit too busy with the base class + work + personal life to really dig any deeper. Key to the group project is to definitely find yourself a good group (Start looking right as the class starts). I feel like I had a good group but we still struggled a with timeline. Grading on the project is pretty lenient, it felt like as long as you put a solid effort forth you got a good grade.

  Rating: 3 / 5Difficulty: 4 / 5Workload: 16 hours / week

• 0OVTZNLDsBAe183oXEBr5g==August 14, 2026summer 2026

  Software Architecture and Design

  I took this in my final semester, and it was by far the WORST CLASS I've taken in the program SADly!!

  As a SWE of 3 years, the content was very outdated: mostly about pedantic diagram syntax (ex. What a black arrow means vs a white arrow in UML. Which no one cares about) and a fire-hose of software design styles (No way you retain these, and -0.001% chance any of them will be referred to as their academic names in real life). I also found the lectures incredibly boring, being much less well-edited than other classes.

  Assignment #1 was a group assignment for creating large UML diagram (~7 related classes) with two sequence diagrams. Did not feel like a useful learning experience at all since UML was the brunt of the work, and as many other reviews say no one cares about those anymore. Assignment 2 and the Exam were also all about the specifics of UML. So mostly impractical, get ready to forget all of this the second you get a SWE job since u wont be needing that info :)

  The final assignment was a full-stack project with a required backend API. That's the useful, practical part: the other half of it is mostly useless diagrams that take time to create and keep up-to-date. It's worth 30% of your grade so don't slack on it! This was the most I've crunched for ANY of the class projects I've taken in OMSCS since my team was new to programming.

  This course had 0 previous coding assignments to help prepare for it. All of my other programming classes had smaller and easier assignments to help familiarize students with code before dumping a project on them. Therefore, if you have non-CS teammates, be prepared to carry them on your back! This pacing of this class is skewed heavily towards the end with this Jumpscare of an assignment.

  Outdated, boring, and stressful even for someone with years of programming experience. Would not call this an easy A due to team RNG for Assignments 1 and 3. Farewell OMSCS and happy the rest of the courses weren't like this one!

  Rating: 1 / 5Difficulty: 4 / 5Workload: 10 hours / week

• 69q9X6ADD/iA7rJ7vsxfNQ==August 14, 2026summer 2026

  Probabilistic Models and Their Applications

  This course was well-taught overall. Professor Song Hee Kim is excellent, uploading lectures that are organized and easy to follow. The homework is challenging, but very manageable with the lectures as a guide. I took this class over the summer, and my main critique is pacing — content got fairly crammed toward the end of the semester. I think spreading things out a bit more earlier on would have made the last few weeks more manageable.

  Rating: 4 / 5Difficulty: 2 / 5Workload: 7 hours / week

• vsibVbdFfYHQ84sN6cGhvw==August 13, 2026summer 2026

  Seminar: Computing in Python

  If you're one of those OMSCS students that have a very limited computing background, then this is the course for you. It teaches you the basics needed for programming in python. This class is basically programming 101 for Gatech and the fact that's it's made available to OMSCS students is amazing.

  I'm one of those students who did not major in CS nor have I worked in the tech industry so my programming is pretty spotty. I was familiar with all the concepts thought in this course, but when it comes to programming, you need to constantly do it for it to stick and this course provides that. For OMSCS, tests and quizzes are not needed, only module assignments so I took this alongside another class and had no issues. Some weeks are busier than others though.

  Seminars are relatively new so wish they were offered when I started OMSCS. Highly recommended if you have no programming experience or would like to brush up before taking actual programming classes.

  Rating: 5 / 5Difficulty: 1 / 5Workload: 5 hours / week

• Iry1J3YMb99kAfRKQq4FPg==August 13, 2026summer 2026

  Computer Animation

  It was very math heavy and all assignments were done in a python web interface but I feel like it was more learning about linear algebra and NumPy than anything else. Pretty boring unless you plan on implementing an animation system within an app like Blender or Maya, I was hopping the projects would be more interactive.

  Rating: 3 / 5Difficulty: 3 / 5Workload: 12 hours / week

• Iry1J3YMb99kAfRKQq4FPg==August 13, 2026summer 2026

  Foundations of Computer Graphics

  The course was well done overall, I did not like the monotone reading from Dr. Wilson's lectures. It was so robotic that I thought it was AI at first. The content itself was interesting and I learned a lot. The use of Java didn't really bother me and this kind of implementation (Ray Tracing & Mesh Manipulations) should really only be done on low level frameworks like Metal and OpenGL but doing it in that would have made it a bigger headache IMO.

  The projects were not easy by any means, but they were a tad boring at times.

  Rating: 4 / 5Difficulty: 3 / 5Workload: 15 hours / week

• ipQJIv6zRXFYvX7BIhf/eA==August 13, 2026summer 2026

  Natural Language Processing

  For context, this was my 5th OMSCS course after AI, ML, Deep Learning, and Intro to Operating Systems. I’m 30 years old and have about 10 years of software development experience.

  Overall, I really liked the course. The lectures by Dr. Riedl and the former GT student were excellent: clear, practical, and easy to follow. Some of the material presented by Meta AI engineers was harder for me to understand, but the course was still very worthwhile.

  The homework assignments were generally manageable, with HW5 being the clear exception. I made a mistake there and did not implement the attention mechanism correctly, but I was still able to recover and do well on the final exam. The midterm and final were hard but manageable if you kept up with the material. The quizzes were fair, and in the second half of the course they felt a bit less challenging, though that may just have been my experience.

  One logistical issue I ran into was creating a Google Colab account as a non-US resident, which caused some friction early on. Overall, I would recommend the course, especially to students who already have some ML/DL background. It is less challenging then ML and DL, but fair, and I learned a lot.

  Rating: 5 / 5Difficulty: 3 / 5Workload: 12 hours / week

• D8C7L4DMq33abLb+tYYaqw==August 13, 2026summer 2026

  Quantum Computing

  I've been interested in quantum mechanics for years and struggled to learn it, and I feel like I made so much progress by studying the foundations taught in this course. While it doesn't teach QM directly, it teaches a lot of the math that supports QM.

  If you've ever wanted a better understanding of linear algebra, tensors, complex numbers, and/or quantum information theory, this is the course for you. Well, the first half of the course is for you. The second half of the course focuses more on benchmarking, error correction, and error mitigation, which I wasn't my cup of tea, but YMMV.

  Good

  • Content: IMO the (first half of) course content is SO interesting. I just feel like I understand how the universe works so much better now.
  • Labs: Fantastic. Labs have an autograder (easy 100), aren't too time-consuming (as long as you read the instructions carefully), and really helped me understand the material. Each was <5 hours.
  • Problem Sets / Knowledge Checks: I found problem sets to be a great way to check my understanding. Combined with the weekly knowledge checks (unlimited attempts, free points), they also served as a great guide for studying for the exams. KCs each took me ~1hr and PSs took 1-3hrs.
  • Paper Reviews: I enjoyed most of the papers. The paper reviews were quite chill, just rigorous enough to push you to engage with the paper, without needing to stress too much about getting full points on the review. Each took 2-3hrs.
  • Lectures: I used the course lectures for the second half of the course and found them sufficient, roughly 1hr/week.
  • Textbook: I only used the textbook for the first half of class content. I also read "Quantum in Pictures" (highly recommend -- its written for children so you can get through it in an afternoon) and a number of chapters from "Picturing Quantum Processes" to get a more visual intuition of the algorithms, but we were never tested that deeply on the algorithms, so it was beyond course scope.
  • TAs: I never went, but we had TA office hours 7 days a week. They were also very responsive to questions.
  • Exams: The midterm and final only 10% of your grade each, and each allowed a double-sided note sheet.

  Bad

  • Lecture videos are not available for download.
  • Lectures: For the first half of the class, lectures were not "bad" per se, but IMO they just aren't comprehensive enough to deeply teach all the math needed in this class. I used Qiskit "Foundations of quantum computing" and Quantum Soar on YouTube as main lectures for this class, and then the class lectures only as supplements if I was having trouble.

  Overall, I'm not sure if I had a very representative experience. I just really enjoyed the material, so I did a lot more than the course required. I spent the 3 weeks leading up to the class learning the math (before this summer, I had never really used imaginary numbers before, so I've come a long way), and even still probably spent maybe ~30 hours/week in the first half, and ~10 hours/week in the second half? I got a high A, but the bar for a low A seems to require far less time/effort.

  Advice: I learned the most by feeding the textbook to ChatGPT and having it give me lots of practice problems. IMO you won't understand the course content without doing lots and lots of practice problems.

  Rating: 5 / 5Difficulty: 4 / 5Workload: 20 hours / week

• w1SovK8k43rCPMHj+/qlzg==August 11, 2026summer 2026

  Digital Marketing

  This may have been the easiest class I've taken, ever. That said, I did used to work at a company that did a lot of digital marketing and I only got a B in the class, but still very easy.

  I didn't do any of the reading or watch any of the lectures. I just did the assignments, got 100% on all of them, took the midterm without studying, got 60%. Then I studied a bit for the final and got 76%, which brought my final grade to just over 80%.

  Maybe this would have been harder if I wasn't already familiar with much of the content having worked in the industry. Regardless, I wanted an easy summer course and I got one.

  Rating: 3 / 5Difficulty: 1 / 5Workload: 1 hours / week

• w1SovK8k43rCPMHj+/qlzg==August 11, 2026summer 2026

  AI, Ethics, and Society

  I actually thought that the content for this course was really interesting, I'm glad that I took it, it really does a good job of pointing out the flaws in AI models and how they can be biased.

  However, the workload was mostly just busy work, it wasn't hard, but just frustrating. The reports and analysis were so tedious. I think I had close to 20 charts and tables in my final report. Just basic, rudimentary stuff that wasn't hard at all, but took time. I'm so tired of JDF.

  Fortunately, it only took a couple of boring hours per week to complete these. Overall, I really liked the content, the actual work has much to be improved, but it didn't take too long, and I got an A in addition to feeling like I took something away from the course.

  Rating: 3 / 5Difficulty: 1 / 5Workload: 2 hours / week

• +P2SNPgxTxx8N5phkJLrpA==August 11, 2026summer 2026

  Reinforcement Learning and Decision Making

  This was a really challenging class. I took it during the summer and directly following Machine Learning (so the foundational concepts were fairly fresh).

  I'd say the older lectures were not super useful. The readings were great - I purchased both the Reinforcement Learning (Sutton and Barto) and Multi-Agent Reinforcement Learning textbooks. I did not complete all of the readings, but those that I did felt really helpful.

  There were AI oral quizzes which took a long time and were not worth alot of points, but basically "free" and reasonably helpful for learning the concepts.

  The projects were really challenging. They were 8 page assignments which did not have as much guardrails as Machine Learning. We were not provided with previous example papers (or given example papers after submission). The experiments took super long to run - which compounds any procrastination issues you may have as a student. There would be times I expected + needed a run to happen overnight, but a memory leak caused a crash and no results. The hyperparameters are very touchy, and simple mistakes in algorithm implementation can cause bad results.

  The final was challenging. I did better than average, but poor performance on some of the papers (with one taking a 20% hit for a late day) led to a C in the class - which means it will not count towards the Machine Learning elective.

  I'd say I still feel like I learned alot, but found the class to be really hard. Not one for the faint of heart or if you have a busy semester. Also - if you're taking towards the end of your Machine Learning specialization, make sure you have alot of time and take care to focus on getting a good grade in the class, so that it counts for your elective. Thankfully, it will not affect my course planning too much, so not a huge deal that it does not count towards my ML elective.

  Absolutely do not take this class without having taken ML first. I thought it was harder than ML and the fact I took ML made RL easier than it would have been otherwise.

  Rating: 4 / 5Difficulty: 5 / 5Workload: 30 hours / week

• vsibVbdFfYHQ84sN6cGhvw==August 11, 2026summer 2026

  Special Topics: Introduction to Computer Law

  This course was a very welcome surprise. The lectures were absolutely excellent and gave you a good sense of how law interacts with technology. The law is ever changing so the instructors try to educate in such a way that doesn't focus on current laws but more of an underlying theme that can be applied to a broad sense when it comes to tech. The material didn't feel out of date and the 2 professors made the material enjoyable.

  The grades were made up of 10% Ed discussion, 20% quizzes, 25% acquisition project, 35% code analysis project, and 10% written assignments.

  Ed and quizzes are self-explanatory. Participate and watch the lectures and you'll do fine. The written assignments are proctored assignments where you try to reword and understand a prompt into layman's (not tech savvy) language. The acquisition project is where you'll compare 2 tech companies and explain how one of them may go about acquiring the other. Kind of basic report type of format. The coding analysis will have you compare 2 IDEs of your choice and talk about possible issues with infringement. This project was excellent and forced you dig deep about each IDE. Its look and feel, its source code, its directories and how they interact, and so on. The rubric was a little unclear so i found myself working on the projects significantly more than other assignments. We had over a month for the acquisition project and a little over 2 weeks for the coding analysis project. I prob spent like 15ish hours (~10 pages) on the acquisition and 30ish (~20 pages) on the coding analysis project. So the coding project was definitely more pressure.

  One of the not as good aspects of the course was the timeline for grading when concerning the projects and written assignments. These were graded weeeeeeks past their due date and even as final grades are some individual assignments were left ungraded. I'm not sure if this is due to the summer schedule or not but just something to prepare for. The TA's were also very responsive so that was also good! Overall, a great class and I'm happy i took it.

  Rating: 5 / 5Difficulty: 2 / 5Workload: 5 hours / week

• exenhSmf5lOOcSoZUrQd+Q==August 11, 2026summer 2026

  Introduction to Cognitive Science

  The lecture videos are easy to watch and the content is interesting though very different from other courses in OMSCS. There are alot of additional readings (sometimes 3 or 4 academic papers) that are tested via weekly quizzes but they only account for a small part of your overall grade. Quizzes consist of several MCP questions you need to complete in a short time but they are open book. The readings can be useful for completing the written exercises. Written exercises are easy to finish and should be simple to get full marks as long as you're up to date on course content and follow the outline correctly. The term project allows you to explore any relevant areas you like and takes up the most time for this course. I really appreciated that TAs provided a sample of exceptional papers from previous semesters so you are able to get a feel for how to successfully structure your project. You also need to complete a video presentation presenting your project paper (5 minute recorded video). TAs were great, very responsive and gave useful feedback. Overall, an enjoyable and manageable course for the summer as long as you are able to stay up to date on the readings.

  Rating: 5 / 5Difficulty: 2 / 5Workload: 10 hours / week

• 4285Zmj8Nvy99KKFozO4iA==August 11, 2026spring 2026

  Machine Learning for Trading

  Strongly not recommend taking this course.

  Much of the course content consists of videos recorded by Professor Balch nearly a decade ago, and the course overall feels quite outdated. My experience with the TAs was also very disappointing. Some of their responses came across as dismissive and condescending, and at times it felt as though students’ questions and concerns were not being taken seriously. I also remember seeing several Reddit posts from other students expressing similar frustrations.

  The assignments and exams were another major disappointment. Instead of focusing on whether students actually understand the core concepts of machine learning for trading, they often seem to emphasize unnecessary details and very specific requirements that add workload without adding much educational value.

  Overall, I regret taking this course though I got an A at last.

  Rating: 2 / 5Difficulty: 3 / 5Workload: 20 hours / week

• 9C/pM7nToo3uf5gsllEqBA==August 11, 2026summer 2026

  High-Performance Computer Architecture

  This is a class with some great lecture content blighted by outdated content (errata filled quizzes), outdated project documents (word documents that are extremely poorly formatted with supplementary Ed FAQs), and slow grading due to the aforementioned project issues.

  I would recommend others to take a different class if you're on the fence and just go through the lectures/book if you're interested in this material as the lecture content is the best part.

  The class started off with quite a few Ed posts from the TA staff (Nolan) with responses to questions pretty quickly in the first 4 weeks. Then, the TAs went completely silent until the end of the semester.

  For the summer, the first month was very heavy in lecture and project content. Then, the rest of the term was very light.

  Most of my time was actually spent on the lectures and notes. The projects were very easy (I have C/C++ experience), but they were decent reinforcement of lecture content ideas. I finished the projects usually within 4 hours of starting.

  Exams were of reasonable difficulty. The midterm was a bit harder than the final due to the time crunch of the first half of the course within a month and the more meticulous problems/calculations.

  Success tips:

  1. Take notes of each module in the lectures since the exams are open note. Anything is fair game.
  2. Practice the long problems presented in the lectures (branch prediction, pipelining, Tomasulo's, ROB)
  3. Use a partner for the projects where it is allowed if only for sanity checking answers. The word documents are easy to screw up (flipping entries) and running the program commands isn't fully dummyproof if you mess up the command itself or forget to change the config file appropriately.
  4. Double check your project submissions to ensure you are submitting your filled out word file. Multiple people accidentally submitted the original, unfilled word file.

  Rating: 3 / 5Difficulty: 2 / 5Workload: 4 hours / week

• 9C/pM7nToo3uf5gsllEqBA==August 11, 2026fall 2025

  Introduction to Graduate Algorithms

  This course was more difficult and laborious than it needed to be for a few reasons:

  1. The tests are rather punitive for any errors in the free response (minimum 8 point deduction)
  2. The professor was unprepared for office hours for the last half of the course, giving incorrect explanations for homeworks
  3. Lack of solutions or proper explanations of solutions for almost all of the assigned problems
  4. Feedback for the free response homework problems were returned the week of the test generally, leaving very little time to fix bad habits or clarify misconceptions in your mind
  5. Extremely long (8+ hr) weekend office hours cannibalized by the same 5 people with dumb questions without published slides
  6. Notations or vocabulary in the test that aren't covered in the preparation material.
  7. 3 hour long tests are a slog in terms of focus and bowel functions (this isn't the only course with this issue)

  Some tips for success:

  1. Join a few study groups and stay with an active one. They are hit or miss
  2. Be prepared for overly verbose multiple selection problems targeting any edge cases on the exams as the first exam can catch people off guard.
  3. Generally, one of the exam free response questions on each exam was similar to the homeworks, so you should at least complete the homeworks as the staff recommends.
  4. Read the staff posted guidelines early and take notes to memorize. Use those guidelines for practicing homeworks, notably the pseudo-proof requirements.
  5. The TA's are there to help, but they aren't going to hold your hand and they aren't going to respond with sunshine and rainbows. Put some thought/effort into the material and ask your study group for help. If something still isn't making sense, post a concise, detailed question with your thoughts and sources.

  Ultimately, my main gripes are the lack of scalability and the lack of support to maximize learning instead of brute force iterations of problems. The long turnaround time for homeworks is a huge issue throughout the semester as they are the primary method of preparation for the exam free response questions. Manual feedback for free response 9questions doesn't scale with the amount of students in OMSCS. The long office hours and lack of problem solutions make the course very time intensive. Most of this is due to the philosophy of the teaching staff that one needs to keep trying the material until the "light bulb turns on".

  Rating: 2 / 5Difficulty: 5 / 5Workload: 20 hours / week

• 2c1btKEcjhtxtHHB8f2rMg==August 10, 2026summer 2026

  AI, Ethics, and Society

  Final Grade: A(96.08%).

  Pros:

  • Very interesting AI/ML Topics
  • Concepts are explained clearly and in a straightforward way.
  • Assignments are manageable and not overly time consuming.
  • Great, easy summer class

  Cons:

  • The midterm did not reflect the material covered in the lectures.
  • Feedback on the midterm was too vague and did not provide a clear explanation.

  Midterm(10% ) - 74.41%: As mentioned in the cons, most of the midterm did not reflect the material covered in the lectures. I spent a lot of time studying for it and was disappointed with the structure of the exam. Additionally, the feedback provided was too vague. It would have been helpful to see the actual questions and explanations for the answers so that we could better understand our mistakes and learn from them.

  Projects (40%) - 96.6%: Overall, the projects were pretty easy. I lost a couple of points for not using physical tables and for not providing screenshots. Make sure to read the project instructions very carefully so you don’t lose points over small details. There were five projects throughout the semester, so make sure you start working on them right away.

  Final project (15%) - 100%: You have the option to work on the project individually or with a team. I chose to work with a team, and it made the project much easier. You can have up to four people in a group, which is what we did, so each person only had to complete about 1/4 of the project. Keep in mind that each step builds on the previous one, so whenever you finish your part, make sure to provide your work and any necessary information to the next teammate.

  Final Exam(10%) 100%: It wasn’t really an exam and was more like a project. You had to write about different topics covered throughout the course. Overall, it was pretty easy and very similar to the projects we completed throughout the semester.

  Class Discussion/Exercises(15%) - 100%: You have to answer a list of questions, share your opinion, and reply to two classmates’ posts. It was pretty easy and only took a few minutes to complete.

  Written Critiques(10%) - 100%: There were only two of these assignments throughout the course. You had to write a short JDF-style written critique that was no more than three pages. Each assignment provided a list of questions that you had to answer in your critique.

  Rating: 5 / 5Difficulty: 2 / 5Workload: 7 hours / week

• 9C/pM7nToo3uf5gsllEqBA==August 10, 2026fall 2025

  Graduate Introduction to Operating Systems

  Lectures are very useful in terms of presentation and content. Projects help with exposure to coding and debugging multithreaded/multiprocess programs, shared memory apis, and GRPC. The first project lacks some detail, but just taking a look at the provided code makes it easy to fill in the gaps.

  Generally, I could finish all projects in two sessions of about four hours, with some edge cases and updated information via Piazza/slack creeping into 4-8 extra hours. Note: I have C/C++ experience already.

  Tests were the most difficult part of the course for me as there's quite a bit of information to understand/memorize.

  Overall, this class has already proven useful to me in my CS job. I wish there were some more modularized classwork coding exercises to explain worst/best practices for the topics covered. Additionally, it would be nice if a review document was posted for each project showing good optimizations and coding practices for learning purposes.

  Rating: 4 / 5Difficulty: 2 / 5Workload: 8 hours / week

• +GZ760wxl6bpNfvPPQ+Ung==August 10, 2026summer 2026

  Quantum Computing

  Something I've noticed over my time with OMSCS is that the professors who teach via whiteboard are engaging and informative (AC, AI, HPCA), while the professors who teach via powerpoint just provide a surface level overview and expect you to learn the practical details yourself (CN, DSI, HDDA).

  CS 7400 - QC - is a powerpoint class. Be prepared to spend around 2 hours of self-study for every 30 minute! lecture if you want to properly understand the topics. At least for the front-loaded first half of the course (9 h/week). The second half is a complete joke (3 h/week). The course is only good enough to familiarize you with the basics of quantum computing and not much of any interesting deep-diving.

  Still, I found the topic interesting and would recommend it if you want a thowaway summer class.

  20% of your grade comes from 4 open-book quizzes and 20% from 2 closed-book Honorlock spyware proctored exams. Recommend you use a LiveUSB to protect your privacy. You will be tested on concepts found only in the textbook.

  Rating: 3 / 5Difficulty: 2 / 5Workload: 6 hours / week

• +GZ760wxl6bpNfvPPQ+Ung==August 10, 2026summer 2026

  Quantum Computing

  Something I've noticed over my time with OMSCS is that the professors who teach via whiteboard are engaging and informative (AC, AI, HPCA), while the professors who teach via powerpoint just provide a surface level overview and expect you to learn the practical details yourself (CN, DSI, HDDA). CS 7400 - QC - is a powerpoint class. Be prepared to spend around 2 hours of self-study for every 30 minute! lecture if you want to properly understand the topics. At least for the front-loaded first half of the course (9 h/week). The second half is a complete joke (3 h/week). The course is only good enough to familiarize you with the basics of quantum computing and not much of any interesting deep-diving. Still, I found the topic interesting and would recommend it if you want a thowaway summer class. 20% of your grade comes from 4 open-book quizzes and 20% from 2 closed-book Honorlock spyware proctored exams. Recommend you use a LiveUSB to protect your privacy. You will be tested on concepts found only in the textbook.

  Rating: 3 / 5Difficulty: 2 / 5Workload: 6 hours / week

• MY2sE7PGyvhZphh0gFqHRw==August 10, 2026summer 2026

  Graduate Introduction to Operating Systems

  The course is rewarding, but entails a lot of overhead time commitment due to poor organization. It is especially important during the summer semester, when the amount of work is the same, but deadlines are shorter.

  I got an A in this course. This was my second course at OMSCS. The first one was Computer Networks. I liked how these two courses paired because at GIOS I touched in practice what I learned in theory in CN. I had no prior experience with C or C++, but had many years of programming in Kotlin and TypeScript.

  Now I will describe GIOS in detail.

  First, as people already said here, some TAs are passive-aggressive, but still mostly helpful.

  Second, all the course content is comprised of video lessons recorded by professor Gavrilovska more than 10 years ago. I appreciate the efforts that she made to express complex ideas in simple terms with beautiful images; however, the videos are rather shallow and effectively useless if you want to prepare for the exam. It is better to only read a book.

  Third, there is no text version of videos. So you watch them, remember them, and that's it. It is impossible to search something and impossible to use these videos to prepare for an exam since videos take a lot of time.

  Fourth, if you have a Mac with an Apple Silicon chip, you are going to have troubles, because the course expects you to work on x86 architecture. A widely-spread recommendation is to just rent for money a cloud VM with x86 architecture. However, there was a guy who prepared a guide how to work locally with Silicon chips without renting a VM. I hope the team will introduce that guide into the course.

  Fifth, the course unnecessarily refers to old papers. I hope the team will find new interesting materials to incorporate. For example, for distributed file systems the course references a paper from 1988 and then we together analyze the workload that was in 1988 for that company for that distributed system.

  To finalize, the course is great and I am very glad that I took it, because I learned a lot, but the time required is unnecessarily extensive.

  Rating: 4 / 5Difficulty: 5 / 5Workload: 22 hours / week

• Mg8VbB4wqnj8Dv27PhxvHQ==August 10, 2026summer 2026

  Digital Marketing

  My background is a BS in CS with about 2.5 YOE as a SWE, and this is my 8th course in OMSCS.

  This was probably one of the easiest classes to get a B in, not just in OMSCS but out of all the college classes I’ve taken over the years. Getting an A takes a bit more effort, though, since the exams make up 60% of the grade. I barely managed to get an A myself.

  As for the course itself, it was interesting and I definitely learned a thing or two, but that’s probably about it. I wouldn’t say there was anything particularly groundbreaking or memorable beyond that.

  Rating: 4 / 5Difficulty: 1 / 5Workload: 5 hours / week

• 3UfbwURUtDngmqbXNrAhQw==August 10, 2026summer 2026

  Data Analytics and Security

  Overall, a very disappointing course.

  The organization is a mess. Modules do not open in order, and some assignments may be due before the module containing the relevant material is even available. This makes the course unnecessarily confusing and difficult to follow.

  The biggest issue is the grading. Despite having detailed rubrics, grading often feels subjective and inconsistent. Points are sometimes deducted for unclear reasons, and questions about grading are rarely answered. For example, I received full credit on one coding assignment, then lost half the points on the next for making a very similar type of change because it was considered “not substantial enough.” At times, the grading feels almost like luck.

  The course content is also disappointing. Despite the title Data Analytics and Security, the lectures mostly feel like a very high-level introduction to data analysis, with surprisingly little meaningful connection to security. The lectures are often vague, disorganized, and not especially useful for the assignments.

  The TAs are also generally unhelpful, which makes the poor organization and inconsistent grading even more frustrating.

  Overall, this course needs a major revamp: better organization, more relevant security-focused content, clearer grading standards, and better support from the teaching staff.

  Rating: 1 / 5Difficulty: 2 / 5Workload: 15 hours / week

• yGQfqHWI/Zm/+Be0x+tsuQ==August 10, 2026summer 2026

  Natural Language Processing

  Lectures: The lectures here are an interesting split. Most of the course lectures are taught by Dr. Mark Riedl, and those lectures are excellent, an easy 10/10. However, about 1/4 of the course lectures (measured by playtime) are taught by various Meta AI guest lecturers, and they vary in quality from 2/10 to 7/10. Dr. Riedl breaks complex topics down into clear and learnable concepts better than any professor I have had in this program so far, and his pacing is excellent. I have a reasonable amount of experience with neural networks, but the lectures were still able to add to my understanding in a meaningful way. As for the Meta lectures, some were reasonable, interesting, and easy to follow, but many were entirely useless, reading directly from the slides in thick accents and adding literally nothing to my learning. Taking all lectures in the course into consideration, the final score I will give is an 8/10.

  Instructors: Dr. Mark Riedl was not super active directly, but he was around from time to time. The TAs were very active and very helpful throughout the course, always responding faster than expected. Many individuals this semester had a lot to say about the rigidity of the course being a bit much. I would say that the staff took cheating prevention a bit too far to the point of inconvenience, but I would rather see this taken a bit too far than see it taken not far enough, so I saw no issue here. In addition, Dr. Riedl actually added a few new lectures during the semester to help keep up with this extremely active and fast-paced field, which I greatly appreciate. Overall, the instructors earn a 9.5/10 in my book.

  Assignments: I loved the assignments in this course. Homeworks were very well defined with unlimited autograder attempts, so straight 100s is very doable. But aside from grading, they are very instructive, and they did an excellent job reinforcing the material. They are Python notebooks with very well-defined steps. They build up ML and NLP intuition and conceptual understanding very well without introducing any overly difficult programming. The quizzes are fair; they are a pretty solid indicator of how much attention you paid during lecture. The quizzes are set up to see if you paid attention, and the exams are set up to see if you understood more deeply. A quiz question might be something like "which of these techniques were discussed," while an exam question might be "explain the pros and cons of this technique when applied in this way". The assignments get a 10/10 from me, with the homeworks and mini project being some of my favorite work in this program so far.

  Difficulty: If you meet all the pre-reqs, its not too bad. The content is a bit demanding, but the lectures break it down very well, and the homeworks are very instructive. I would heavily recommend taking good notes from every lecture (including small details), as studying this content is very important for the exams.

  Time I Spent Weekly Most weeks, I spent around 10-12 hours watching lectures, taking the quizzes, and completing the assignments. However, I took this as a summer course, so I imagine this workload is less during a regular semester. This was a pretty manageable course for the most part aside from the home stretch, where I spent more than double this amount of time in the closing weeks to work on the final project and study for the final exam.

  Rating: 5 / 5Difficulty: 3 / 5Workload: 11 hours / week

• ygM6gVpSGujRmaIHCF7+8w==August 10, 2026summer 2026

  Deep Learning

  Overall great course probably my favorite so far from 5 completed (ML, RL, AI4R, SDP, DL). Got an A by a fraction of a percentage point.

  This class was very good because it forced me to learn concepts that I knew at a high level in great depth and detail. For example, I 'knew' gradient descent vaguely but it's a whole new world of understanding after doing entire passes of it by hand. Same goes with many other core deep learning concepts such as each of the critical components of a basic feedforward NN, then CNNs, RNNs and attention, before the class turns towards surveying many methods in a bit less detail. TBH, I learned so much.

  What was also great was that every single part of the class feels like you are not wasting your time, you know why you are doing what you are doing, what the learning objective is, and it is no nonsense. To borrow a previous students phrasing a "practical pedagogical style that is pleasant for the summer" (paraphrase), despite the significant workload. The straightforward grading and the lack of tedious long form reports make this class feel better bang for your buck than ML and RL.

  The quizzes were definitely tough, and I studied a lot, but still had a tough time on them. They are tougher as the class goes on, but become less math heavy.

  Group project I found a group early and we got it done and it was not too bad, and they graded leniently.

  Overall though, it was still a ton of work, and i had to really refresh my calc 1/2 and linear algebra knowledge at the start of the course. I studied extensively for the quizzes and still struggled a bit. The assignments took quite a while but if you got started early enough you would be okay, considering the test suites they provide tell you if you are on the right track. Lectures solid, as most agree, the second half facebook lectures are worse. Whenever there was a section I was struggling with, I would watch Justin Johnsons university of michigan lectures as supplemental material to just get a diff perspective on it, and that helped.

  If you are interested in deep learning / neural networks, definitely take the class.

  Rating: 5 / 5Difficulty: 4 / 5Workload: 14 hours / week

• NR73u2Tj3x0c2yTEK/kJAw==August 10, 2026summer 2026

  Machine Learning

  I took the course having a solid prior knowledge of ML. I got a very high A, with all my reports' grades above 97 (before reviewer response) and a final exam at 87%. Despite having both a good theoretical and practical knowledge of ML, I have learned a lot from this course, and it's now among my favourite OMSCS courses I've taken! This term the thresholds were A>82.52, B>70.71, C>58.9, D>47.09; the final exam average was 77%.

  Pros:

  • It felt like a proper graduate course, with very-little-to-no busy-work and with purposeful and instructive assignments, challenging at the right level.
  • Stellar instructor and TAs, very involved and supportive, and clearly caring a lot about creating a good environment for learning.
  • Very clear (albeit lengthy) rubric instructions and expectations for each report.
  • I really liked the hypotheses-oriented framework expected for the reports, and the focus on analysis rather than on coding.

  Cons:

  • It can be a very time-consuming class, especially for the report writing.

  Rating: 5 / 5Difficulty: 4 / 5Workload: 20 hours / week

• 9AIBpLdKh5gudR0j51RJBQ==August 10, 2026summer 2026

  Machine Learning

  The good: if you have prior ML experience, just skip all videos/lectures and focus on the reports. The first report took longer (30 hours), but subsequent ones can be done in 15-20 hours each due to familiarity and re-use of code/LaTex snippets.

  The bad: the summer introduced peer responses on the reports and some of the posts or comments were just atrocious and really farming for points with no value add. Ed participation also had a similar dynamic, one group of genuine discussions and questions, then contrived posts about what someone learned or some very hand-wavey qualitative CS6603 'realization' post about some basic ML concept to farm the point. I guess to be expected when you have students trying to farm just to pass, but it takes away from the course to see this behavior.

  Time commitment: reports: 80 hours, quizzes: 5 hours, responses: 3 hours, final prep: 5 hours. Average ~8 hours per week over summer.

  Rating: 3 / 5Difficulty: 3 / 5Workload: 8 hours / week

• zIZzXHYQ2jFLtOKoXRineg==August 10, 2026summer 2026

  Machine Learning

  Managed to get a high A in this class in the summer, which I'm extremely happy about. I had read many reviews before taking this class, so I knew what to expect. This is one of those classes where you are thrown into the deep end and expected to produce in-depth reports with barely any teaching. Therefore, if you already know how ML projects work (EDA, preprocessing, training, tuning, etc.), then it provides a good opportunity to get exposure to that. If you don't have experience with those main ML concepts, then I would NOT recommend this course.

  This class required by far the most work out of the 7 classes I've taken so far, although RL was comparable. I chose to pretty much skip all lectures and additional readings to instead focus fully on the reports. This did degrade the learning a little bit, but also allowed more personal time (still watched world cup, went to disneyworld, etc.).

  This is also the first class I've taken which allows use of LLMs, and I'm not sure the class would be possible without it. P2 for example requires a full Pytorch model to be built with custom optimizers and stuff -- I took DL and still can't code that! I relied heavily on the LLMs for code/latex generation, understanding the project requirements, and helping translate insights into analysis. I understand the critique of allowing this, but I was still learning a lot despite this. Plus, this has become the standard in the industry.

  The quizzes were stupid and pointless. You can see the questions in between attempts so the strategy is to figure them out in between attempts. Not sure if this was intentional or not.

  The reports were brutal and very time consuming, but it felt really good to see the final product once each report was finished. It also allowed the opportunity to get down and dirty with analyzing certain aspects of models, algorithms, and techniques. I honestly learned a lot about different inductive biases, optimizers, dimensionality reduction methods, clustering, and randomized optimization algorithms. The key for the reports is to be very clear about how you set up your experiments, and to try to connect every result with a "why". I feel a lot more comfortable performing graduate-level research and analysis now.

  ADVICE TO SUCCEED:

  • know sklearn, some pytorch, latex, and ML project structure beforehand
  • skip lectures
  • get an LLM to summarize any additional readings, or just skip
  • look at the "good" report examples they provide for reference
  • don't be afraid to rely heavily on LLMs for code generation, just make sure you understand everything to be able to draw insights and analyze in your own words

  Rating: 4 / 5Difficulty: 5 / 5Workload: 20 hours / week

• VZuizvJ61QQ8b9TKcMFtTQ==August 9, 2026summer 2026

  Quantum Computing

  I took this course to take a break from the ML spec to see what else was out there. It was quite math heavy at first, but very manageable. After the midterm, it went very broad. I'm glad I took it to dip my toes into QC, but I also know now that I probably don't want to go into QC. The field is just way to nascent for me. To that end, I wasn't a fan of the class, but that's more of a me thing rather than the class. Take this class if you're quantum curious

  Rating: 3 / 5Difficulty: 4 / 5Workload: 1 hours / week

• Y9eyEGsEr80iHDA/JcdJUA==August 9, 2026fall 2025

  Distributed Computing

  I've taken other OMSCS courses such as AOS, DL, ML, and AI. DC was my favorite since it is a subject I'm least familiar with and most challenging assignments. But there are definitely some areas to improve.

  Lecture material: the first half of the lectures covers foundations and theory of distributed computing and the latter half covers implementation and applications of distributed systems. I thought all lecture content was interesting and you can read/skim the assigned papers to get a better understanding. I learned a lot going through the material and studying for the exam. But I would say understanding of the lecture material isn't evaluated in a super rigorous way. There are no proofs or problem sets like you might encounter in other DS courses. The exam questions could be better checked for clarity or underlying assumptions. The instructors understandably hide the questions after you take the test. As an unfortunate result, they probably don’t get much feedback on specific questions to either correct, improve, or clarify. Labs: could be way better paced. Lab 0 is trivial and Lab 1 is also easy. Could knock out both in 1 week. Lab 2-4 are much more challenging. Lab 3 (Paxos) is especially a time blackhole – spent close to 150 hours and got around 90%. By the time Lab 4 came, I burnt out and settled for a lower score. Adding another lab (like Raft) could be interesting or it could just be more burn out, but the schedule backloading has to be fixed first.

  I'm not a professional software engineer but do have significant coding experience. I would say if you’re familiar with object-oriented programming, you shouldn’t really have any issues. Using the recommended Java IDE + Lombok made coding extremely pleasant. The actual challenge with Lab 2-4 was figuring out all things that could go wrong when implementing your protocols, reviewing the sequences of events, and spending time using the visual debugging and modeling tools to fix errors. As another reviewer put it, any little thing that could possibly go wrong almost certainly will when exploring the large state graph.

  Overall, I gained an appreciation for the systems underlying many large modern applications.

  Rating: 4 / 5Difficulty: 4 / 5Workload: 30 hours / week

• 7g/spglPuPsB5FGSd3EWig==August 8, 2026summer 2026

  Computational Data Analysis: Learning, Mining, and Computation

  I knew people who took this course said this is one of the best courses in the program. I now agree. The lecture is great. Each homework is hard but TA gives so much support with office hour being hosted everyday (recorded) guiding students. Also the grading was so generous. I consistently got 90+% even though I didn't get the correct answer because the effort was made. It seems like they want to reward the effort to encourage and maximize learning. I appreciated it.

  The course material goes over various ML algorithms in both mathematical details as well as coding implementation. After this course, you will never be scared by math. I learned so much. Overall the quality of lecture, TAs, homework assignments were all top notch. I'm glad I took this course.

  Rating: 5 / 5Difficulty: 3 / 5Workload: 10 hours / week

• +l+4cqss4ZJbaM6nhPhIkg==August 8, 2026spring 2026

  Data Mining and Statistical Learning

  The lecture quality was bad and didn't really relate to the homework and project (and final exam) they tested students on. The lecture video had the professor going over lots of math very fast without explaining details. Like many of my class mates, I stopped watching the lecture in the 2nd week. Also, while I appreciated the homework assignments trying to get students to experience the end-to-end ML analysis, the grading was so random and subjective with whichever TA gets assigned to you, so we never know if we are gonna get 70% or 100% (kinda like 6501 peer review). To me, this is not a great course because it's relatively easy material but not easy to secure an A. It's similar to 6501 in that sense also.

  This is a world of difference compared to 6740 whose lecture was really thorough and related to the details of each homework assignments, and its grading was generous, and letter grade A is practically guaranteed as long as you make a decent effort.

  Rating: 1 / 5Difficulty: 4 / 5Workload: 12 hours / week

• Weoup0Z707a6VPLR+vIhqw==August 8, 2026summer 2026

  Statistical Modeling and Regression Analysis

  Great course. I thought I knew regression model well from 6501 but this course really goes deeper into the statistical properties and modeling assumptions of regression models. How to check and handle correlation among input features, how to handle differently scaled variance of residuals, in what situation is the statistical significance inflated vs deflated, how to quantify the tradeoff between bias and variance, so on.

  A lot of these things I knew conceptually but now I feel confident in knowing them mathematically. So I can answer if I get quizzed in job interviews. This course really gave me the foundation I was looking for.

  The assignments were a mix of quiz, coding, proctored exam, and a group project. Overall it was not too difficult but obviously busy because of the compressed schedule of summer semester. Group project can be annoying but the grading was lenient so it turned out fine.

  Rating: 5 / 5Difficulty: 2 / 5Workload: 3 hours / week

• 2Q7nfDOZbID+NXdsSK4cPQ==August 8, 2026summer 2026

  Human-Computer Interaction

  I wish I had only good things to say about this class because I really enjoyed course content and its application, but summer semester was a TON of work. I did not know this deep of an understanding was required to understand HCI and a lot of the concepts in this class can be applied to many aspects of work and life. However, I did have quite a few issues with this class.

  Pros:

  • The Joyner structure was very well defined at the start of the class, not too difficult to look/work ahead.
  • Lecture were great. Dr. Joyner did an awesome in creating these. Even though they are a little outdated, he did add some videos for modern-day/AI application.
  • Individual and team projects were a lot of work but rewarding. Cons:
  • Content was hardly modified for summer semester, there were some weeks in the middle of the class where I was putting over 20 hours of work.
  • Some inconsistencies in course instruction between this semester from previous semesters.
  • Quizzes were a little difficult as we went in not really knowing what to study.
  • I guess this is a pro and a con but wouldn’t say it’s really worth studying too much for the exams. Spend more time making a detailed study guide and be ready to search through the readings for the end part of the exam.
  • Probably the biggies issue for the class - grading was quite inconsistent among different TA’s/students for things like homework, projects, and quizzes.

  I expect to receive an A in the course and don’t think any part of the class was too conceptually difficult, but if I were to take it again I prob wouldn’t in a shortened semester.

  Rating: 3 / 5Difficulty: 3 / 5Workload: 14 hours / week

• HYfqosSVeZIk6hC4Zx6ARg==August 7, 2026summer 2026

  Machine Learning

  CS 7641 is a core OMSCS course that covers a wide range of topics including Supervised Learning, Unsupervised Learning, RL, and Randomized Optimization. The summer schedule makes it fast-paced, so good time management is crucial.

  Key Takeaways & Advice:

  Reports over Code: The grade is heavily determined by how well you analyze, visualize, and explain your experiments rather than just getting high test accuracy. Make sure to address the "why" behind the data patterns.

  Pacing: In a summer term, deadlines come quickly. Start the analysis and writing early as generating experiments and writing thorough reports takes time.

  Exams & Material: Lectures give a solid high-level intuition, but reviewing problem sets and participating in Ed discussions is essential for exam preparation.

  Overall, it's a challenging but rewarding course if you enjoy hands-on experimental analysis and machine learning theory.

  Rating: 5 / 5Difficulty: 5 / 5Workload: 12 hours / week

• CzL137OPlAU4vICQteiYpQ==August 7, 2026spring 2026

  Natural Language Processing

  Pros:

  • Dr. Reidl's lectures were fantastic. I work in the AI/NLP space and the knowledge I've acquired through truly watching the lectures and taking notes has been significant.
  • Homework assignments were not an unreasonable amount of work. I kind of wish there was less pre-written code for us, and we were able to code more, but from a workload perspective, these homework assignments reasonable.
  • The exams (midterm and final) are extremely fair for students who take the time to study and understand the material. The short answer questions allow for demonstration of knowledge of the material, making it easy to get partial credit even if your answer isn't perfect.
  • The TAs were highly responsive to feedback, and if they felt there was a grading mistake or unfairly worded question, they would quickly make amends.

  Cons:

  • The second half of the course was dominated by Meta AI lectures, which were a significant dropoff in quality from Dr. Reidl's lectures. The Meta AI engineers giving the lectures seemed unprofessional and disinterested and spent most of the lectures reading off slides rather than actually explaining the material.
  • I don't mind closed book quizzes, but I often felt the design of the quizzes can be unforgiving. For example, being asked only 2 questions on quizzes spanning 10+ lectures of material made it easy to miss a significant percentage of points even while knowing most of the material.

  Tips for success:

  • Start assignments early.
  • Really watch the lectures and take paper notes.
  • Print out diagrams of complex architectures (LSTMs or Transformers) and annotate them as Dr. Reidl goes through the lectures.
  • Have a partner help you study for tests by asking you questions based off your lecture notes.

  Rating: 4 / 5Difficulty: 3 / 5Workload: 9 hours / week

• 4OfsRp+zID+gMr0n4eG0Xw==August 7, 2026summer 2026

  Artificial Intelligence Techniques for Robotics

  I paired this course with another relatively easy course, and the workload was very manageable. Even during the summer semester, I was able to finish every programming project before the midpoint of the course. I earned full marks on all of them except the Path Search project, where I received a 98. The projects were challenging enough to reinforce the concepts but never felt unreasonably difficult. If you like planning ahead, it's definitely possible to front-load much of the work. I also found the course materials to be clear and sufficient for completing the projects successfully.

  Before taking the course, I had heard many stories about how difficult it was to tune. Honestly, I don't think the reputation is entirely justified. Since the programming projects are autograded, you always know exactly where you stand. You can continue submitting to Gradescope until you're satisfied with your score, which removes a lot of the uncertainty. Another thing I appreciated was the quick turnaround for grades, and most were released the next day, so there was no long wait for feedback.

  Personally, I never attended office hours with the professor or TAs because the lecture notes and supporting materials were enough for me to complete the assignments. That said, I think it would be even more helpful if they provided walkthrough videos from previous semesters' TAs, as those could make some of the more challenging concepts easier to understand.

  The TAs were excellent. They were responsive, supportive, and even brought a good sense of humor to the course, which made the overall experience more enjoyable.

  The exams were fair and closely aligned with the course material. As long as you genuinely understand the concepts rather than simply memorizing solutions, it's very possible to perform well. In my case, after my midterm score, I only needed about 30% on the final exam to secure an A, which made the end of the semester much less stressful.

  Overall, I thought this was an excellent course. It strikes a nice balance between teaching important computer science concepts and maintaining a reasonable workload. I learned a lot without ever feeling overwhelmed, and I genuinely enjoyed the experience.

  Rating: 5 / 5Difficulty: 2 / 5Workload: 12 hours / week

• 4OfsRp+zID+gMr0n4eG0Xw==August 7, 2026summer 2026

  Introduction to Cognitive Science

  This was such an enjoyable and manageable course. Each week consisted of assigned readings and lecture materials, followed by a short quiz based on the content. In addition, there were six individual exercises throughout the semester. I found these exercises quite straightforward as long as I kept up with the lectures and stayed within the required word limit.

  The course also included a semester-long project, which I highly recommend starting early because it takes a significant amount of time. I chose a topic closely related to my work, which made the research and report writing much more engaging and rewarding. I think selecting a topic that genuinely interests you makes a huge difference in your overall experience with the project.

  One aspect I really appreciated was the teaching team's responsiveness. Assignments were consistently graded within about 10 days, and the TAs were supportive and did their best to help students. They also listened to student feedback and made genuine efforts to address concerns whenever possible.

  My only complain is that the assignment rubrics were not disclosed. While the grading was generally fair and even quite generous, having access to the grading criteria beforehand would have made it easier to understand expectations and feel more confident about the assignments.

  Overall, I highly recommend this course. The topics were interesting and relevant without being overwhelming, and I genuinely enjoyed the learning experience. I'm very glad I chose to take it.

  Rating: 3 / 5Difficulty: 1 / 5Workload: 5 hours / week

• YssJYlRne4B58yLm6opesQ==August 7, 2026summer 2026

  Graduate Introduction to Operating Systems

  I found this to be a valuable but challenging course. I think I passed, but do have some regrets and suggestions.

  First - avoid taking GIOS in the summer on an accelerated schedule, unless you either (1) have recently taken an operating systems class, or (2) already have a good grasp of parallelism, memory management, inter-process communication, and other topics covered in this course. If this is truly your first real introduction to these subjects, then you should save GIOS for Spring or Fall. While this may boil down to a personal skill issue, I found the lectures and readings somewhat difficult to follow, and wish I had chosen a longer semester to give the material more time to sink in.

  Second - do not take this course unless you either (1) have recent experience with C, or (2) are willing to add 10 hours to your weekly time commitment. The projects assume familiarity with C/C++, and will not gently introduce you to these languages if you lack basic proficiency. Consider taking the C language seminar first.

  Rating: 5 / 5Difficulty: 5 / 5Workload: 30 hours / week

• xyvnNusnYMS7+5rptxb4iw==August 7, 2026summer 2026

  Human-Computer Interaction

  Dr. Joyner does an incredible job teaching this course. His lectures are, hands down, some of the best I've seen throughout my higher-education experience. For a class of 400+ students, I still got the impression that he really cares about us and our success. I took this course during the summer semester, which probably could've come with a warning: this class is busy regardless of the semester, but over the summer it gets compressed, making it feel even busier. There were 4 written homework assignments, 4 quizzes, 2 tests, 2 projects (an individual and then a group), and a smattering of surveys. The quizzes and tests are proctored via Honorlock. The quizzes are closed-note, closed-book, and the tests are open-note, open-book. The group project is almost a copy-and-paste of the individual project, just a little longer and expects higher-quality results. For me, the quizzes came with the biggest barrier. I really struggled with the first one: they tell you the format, but once I did it, I realized I had no clue walking in. Each quiz is 5 free-response questions that test you on specific terms and definitions. None of the coursework is hard but does demand a good chunk of time. If you put the time in, you'll do well. If you think you might want to pursue the HCI specialization, I'd highly encourage taking this course early because you'll learn a lot regardless, and it'll give you a great idea if it's something you want to pursue further.

  Rating: 5 / 5Difficulty: 3 / 5Workload: 15 hours / week

• kNRVENr7tCxqlprILmWMUg==August 7, 2026summer 2026

  Introduction to Graduate Algorithms

  Overall, the class and its TAs aren't quite as bad as it's hyped up to be. Pay attention to Ed, keep up with the material, and put in the effort to practice beyond what's assigned (use AI to help), and it's very manageable. First time taking it and I finished with a B despite bombing Exam 1.

  Exam 1 was by far the hardest and feels like the 'weed-out' exam. I went into it thinking I had a solid understanding of the material and bombed it. Exam 2 was more manageable but still required a lot of preparation. Exam 3 was the most straightforward and the free-response questions were essentially simple variations of the two NP-completeness homework assignments given by the class. The multiple-choice questions across the exams are mostly gimmes if you actually watch and understand the lectures and do the content quizzes.

  My biggest recommendation is to do more practice problems than they provide. AI was extremely useful for generating problems similar to the homework/practice problems and creating variations covering concepts or reductions that weren't explicitly assigned. Don't just memorize the provided solutions, practice applying the same concepts to unfamiliar problems as that was the difference for me after Exam 1.

  Rating: 5 / 5Difficulty: 3 / 5Workload: 15 hours / week

• D8C7L4DMq33abLb+tYYaqw==August 7, 2026summer 2026

  High-Performance Computer Architecture

  Overall, great (but not perfect) class. Highly recommend to anyone interested in computer architecture.

  • Lectures: Lectures are available for free (on YouTube too!). Quality is great. They are from 10+ years ago, so a lot of what was considered "high performance" is now just "regular" computer architecture. But I haven't found any better opencourseware lectures, so 10/10 recommend the lectures to anyone interested in computer architecture. Its a lot of material, and I probably watched each lecture an average of 3 times each (1-2 times originally, and then another 1-2 times while studying for exams).
  • Projects: Project 0 I thought was really cool. I learned how to use Ghidra to decompile a program and figure out what its doing. Projects 1-3 involve interacting with a computer simulator. It was cool the first time to learn "this is how a computer architecture simulator works", but then felt repetitive. Maybe I rushed through them, but I don't feel like I learned much from them beyond that. Each project was 10-20 hours. The TAs gave a fairly comprehensive FAQ and let us benchmark with each other, and that was extremely helpful.
  • Book: I read a bit of the book, but honestly there's so much content in the lectures already.
  • Exams: The exams were open (physical) note, so I thought they were more than fair. This is the only class I've taken so far that lets you see what you got wrong on the midterm, which was so valuable. It turns out, I knew all the material, I just kept making silly mistakes. My advice for exams is to practice questions (e.g. from the practice exams) a few times before the exam to minimize the chances of making small errors. The cutoff for an A is 90, so small mistakes add up quickly.
  • TAs: As the previous review said, they were extremely slow at grading, not very responsive, and even ghosted us for a long time. Not sure what happened. It's a shame, because I think some back-and-forth with TAs would have really aided in learning from the projects.
  • In retrospect, I wish I did the problem sets. They probably would've helped me learn the material a lot more than the projects.
  • Grade: I'm at the cusp between A and B. Fingers crossed.

  Rating: 4 / 5Difficulty: 3 / 5Workload: 20 hours / week

• Y//78ivuAYK34qoqqIUJsA==August 6, 2026summer 2026

  Computer Graphics in AI Era

  I truly and deeply loved this course, especially since it somewhat converges with 3D Computer Vision through differentiable and inverse rendering.

  The Content

  This course covers a lot of material: ray tracing, signed distance functions, volume rendering, radiance fields, Gaussian splatting, differentiable and inverse rendering, physics simulation, and generative models, from classical, differentiable and neural approaches. Some of these topics are so complex that one only touches the intuition and surface details (NeRF and 3D Gaussian Splatting), and implements lighter versions of the techniques.

  The Lectures

  There are too many of them, and sometimes a bit too lengthy. Some weeks require 4 hours of lectures, and some others even 7. Just for lectures. This is not necessarily a complaint, since they are very detailed, very well explained, and very thorough, while still being entertaining to watch. On the bad side, you need to actually be interested in the subject to watch them all since the lectures cover a lot more material than the assignments, and if you're the kind of person who only studies what's covered in the assignments you'll end up watching/learning less than half the content.

  The Assignments

  There are 8 assignments divided in 6 topics. Each assignment is worth 8% of the final grade (64% total), and basically anyone can work on them whether they're enrolled or not since the assignments' site and their repo are publicly available. You can also frontload since all the assignments are released from day one; you'll only need to keep track of submission dates.

  • RAY TRACING is about implementing the fundamental parts of a Whitted ray tracer. For people who took CG this feels like a review while still adding bits not covered in that course. For other people, this one might be the most time-consuming assignment since it asks you to implement many algorithms not very complex but not necessarily trivial.

  • SIGNED DISTANCE FUNCTIONS is about implementing implicit SDF primitives and simple CSG operations. This one is actually very easy and straightforward. The most interesting part is implementing the sphere tracing loop; something not required in the CG course.

  • NEURAL IMPLICIT SURFACES is the complement of the above, where one trains a SIREN network that learns to represent SDFs from input point clouds (actually, sampled meshes). The main challenge comes from accurately translating formulas to PyTorch.

  • VOLUMETRIC RENDERING is about implementing ray marching on a couple of density-based pre-defined volumes. This topic is also covered in the CG course but only at the lecture level; here one actually implements it. The hardest part is not the actual assignment but all the theory behind light, photometry and radiometry, which is quite dense.

  • NEURAL RADIANCE FIELDS is the complement of the above, where one implements and trains a tiny NeRF that fits in a GLSL fragment shader. Here you are given A LOT of starter code since the network is very complex to create from end to end.

  • 2D GAUSSIAN SPLATTING is a deeply simplified exercise on Gaussian Splatting that kinda feels like cheating, since full 3D Gaussian Splatting is also fairly intimidating. This whole assignment feels like implementing only one of the many steps in 3DGS. Still, the actual assignment is kinda fun and visually pleasing.

  • XPBD is about classical physics simulation based on particle positions where one must define a set of constraints and their gradients. After watching the relevant lectures I think I finished this one in around half an hour.

  • DIFFUSION MODELS is about implementing DDPM from scratch in PyTorch and GLSL. Really easy if one took DL before, although that's not necessary.

  The particularly interesting part of the assignments is the CREATIVE EXPRESSION sections that all assignments have, where we can do whatever we want that builds on top of everything and anything covered so far. Even if the core assignments could take you less than 2 hours each, if you're really interested in the course, this could take you over 10-15 hours just for 1 or 2 points this is worth. I'd argue that this is the most important part of the assignments and the course: the opportunity to go above and beyond mostly for ourselves (and optionally for others to see).

  The Midterm

  Really easy; as easy as the problem set. Here you'll solve a few exercises by hand, and step by step, and you'll upload your handwritten results. This might be the weakest aspect of the course; more of an excuse to review the material covered so far (which is still not bad).

  The Final Project

  As lovely as the Creative Expression sections. You can do whatever you want, creative or technical, as long as you cover at least one topic from the course (except ray tracing; that's basically a review and not the core of the course). Here you'll notice just how creative or inventive some people might be. This term there were many beautiful projects and at least one really awesome that displayed the deepest dedication.

  Overall

  All in all, I'm still undecided on whether this might be my favorite OMSCS course or not, only competing with RAIT, which was a really fun course, even if somewhat easier than this one. Still an amazing and well-crafted course that I'll love reviewing from time to time just for the sake of watching these awesome lectures once again.

  Deeply recommend it. You don't need to take Deep Learning of Computer Graphics to follow along but they're quite complementary with not much overlap.

  Rating: 5 / 5Difficulty: 2 / 5Workload: 8 hours / week

• XxopPVjL0mU+bPEGTNxVbw==August 6, 2026summer 2026

  Knowledge-Based AI

  This is my ninth course and by far the absolute WORST class I've taken in this program. Here's a breakdown of why that is:

  1. The core class project. A full 40% of the course is based on completing problems from the ARC-AGI foundation. You will also find that none of the lectures apply to actually completing this project. A question that came up rather frequently during this class between my peers was "how do I actually do this?" Now, academic challenge is expected in any grad level courses but it's another thing entirely if the contents of the lecture don't come anywhere near what you are expected to do. To code up the "AI" agent, you basically are solving all problems for it. Everyone that I've talked with in this class had to basically hand write an agent that will solve the 48 problems it'll be expected to solve for the project. That's not AI, that's me as a person solving leetcode problems. Another thing, you'll spend hours upon hours of your life coding up the agent just for that to only be worth 50% of your grade. Being a traditional Joyner class, you'll have to write a meaningless paper and hope that the TA's aren't feeling too harsh when they grade it. The worst part about this is that the ARC-AGI foundation almost prides itself in being exceptionally difficult for AI systems to solve. Click the link here to see how flagship models perform against these problems https://arcprize.org/leaderboard.

  2. Labs. At the start of the semester, there were 5 AI-based labs that had us interacting with various AI models to learn how different ones behave. This sounds good in principle, but what really happened was we were used as a way to gather data about how the AI systems were performing. The AI systems were so poorly designed that the entire lab assignments had to get scraped and replaced with the normal homework assignments that previous semesters had. If you want to pay almost $1,000 in tuition fees just to act as a test subject for an LLM you don't care about, then this is the class you'll want to take. And don't forget! these labs also come with their LaTeX formatted papers you are required to submit.

  3. Material. None of the things that you'll be learning about are things novel or interesting in nature. This class will take exceptionally straight-forward concepts and somehow turn it into hours long lectures, where the material is so dull and dry that you'll be fighting to stay focused past the first few lecture videos. The two exams are worth 10% of your grade each, and are each 22 five-answer multi-correct multiple-choice questions long. Instead of truly testing you about learning the course content, each question and is a trick question meant to test your English skills more than they will challenge your AI knowledge base. I've watched every lecture video and gotten basically nothing out of this class.

  With the new curriculum change, there are so many better options for you to choose from if you're in the AI specialization. Some classes are hard, but you'll gain a lot from them. Some classes are easy, and you likely won't learn too much in the process. Very rarely, are classes hard, and make you feel like you spent the last semester learning absolutely nothing. This class is one of those very rare exceptions.

  If you want a deeper dive of what taking this class will feel like, see this reddit thread: https://www.reddit.com/r/OMSCS/comments/1udptbh/kbai_might_be_the_lowest_quality_class_ive_taken/?share_id=gAyWWEsUlFP5Vu3tYGPKT&utm_content=1&utm_medium=ios_app&utm_name=ioscss&utm_source=share&utm_term=1

  Rating: 1 / 5Difficulty: 1 / 5Workload: 30 hours / week

• ydNHMgD3s2MQp4YFNtmYWA==August 6, 2026summer 2026

  Special Topics: Global Entrepreneurship

  If it wasn't for the very annoying customer interviews, I'd give this course a 5/5. It is not a challenge in the slightest, but I found it interesting and the lectures are some of the most well-done I've seen so far. Great summer class if you want a breather. Also a fantastic course to pair with another if you're looking to double up (I did not).

  Rating: 4 / 5Difficulty: 1 / 5Workload: 4 hours / week

• Fzg/V2rNZavG9lC/ZedJMQ==August 5, 2026summer 2026

  Introduction to Graduate Algorithms

  Pretty reasonable course! For background, I did a coding bootcamp ~5 years ago and have been working full stack since then. Did humanities in undergrad so never took an algorithms course except the MOOC as prep for OMSCS a few years ago.

  My approach was to read the relevant chapters in the book for each lesson, then go through the lectures. Often the lectures were watered-down retellings of the book chapters - definitely not my favorite lectures and the professor tends to over-simplify things, but they're not bad as a recap after reading.

  Then I did the recommended practice problems and looked up solutions online to check my work. That's literally it - read the book, watch the lectures, and spend time on practice problems. I'm ending with a low A or high B depending on my Exam 3 MCQ performance.

  I think the difficulty of this course is overblown. If I can do this with a humanities background, so can you! Just have to pay close attention to the guidance posts in Ed and follow their desired format for each problem type. You have to ignore some of the whiny reviews on here -for Exams 2 and 3, they practically gave us altered versions of some of the easiest practice problems for their respective chapters in the book.

  They aren't trying to trick you and they aren't trying to fail you. I'd suggest coming into this course ready to plenty of practice problems and you might be pleasantly surprised by how little work is actually required.

  Rating: 4 / 5Difficulty: 3 / 5Workload: 8 hours / week

• JSYEBKrCp/OkgYNve7bHkw==August 4, 2026summer 2026

  Digital Marketing

  Background:

  • eng UG; nonCS
  • 10 y exp in industry: test and devops
  • 9th course; have taken GIOS, SAT, AI4R, CN, AOS, IIS, VGD, CG

  Time:

  • 5 Major cases: 20h
  • 10 weekly mini cases: 12h
  • midterm, incl study time: 3h
  • final, incl study time: 4h

  Experience:

  Very easy class to take a break over the summer before heading into GA. Material is ok, slightly outdated and mostly common sense. Good way to take the summer off.

  Rating: 4 / 5Difficulty: 1 / 5Workload: 3 hours / week

• w8O28GZbi4QsOKRokvPQ3w==August 4, 2026summer 2026

  Knowledge-Based AI

  This class was fine. I switched to the AI track after the debacle of Spring 2026 GA Exam 1 (75%+ of the class failed), and so I'm taking this as one of my last courses in the program.

  It's a standard Joyner course, so there's a lot of writing, a lot of smaller assignments, and a bigger assignment that you work on over the course of the semester. The lectures talk about raven's matrices, but the actual project is ARC-AGI.

  Everything is a little loosey-goosey with how you're expected to solve the problems, and so some of the techniques for KBAI don't click until later in the semester. There's a lot of writing, but that's to be expected.

  So what this means is - if you've taken DL or ML or some of the more meaty classes, this is going to feel like a lot of busy work. If you've not taken a lot of classes, this will force you to reflect and think through problems in a way you've likely not done for a decade or more.

  Overall, it's fine - not a recommend or unrecommend. Better if you take it earlier in the cycle than later.

  Rating: 3 / 5Difficulty: 3 / 5Workload: 15 hours / week

• 0qNdCyeDuSTnuVthrbLt9g==August 4, 2026summer 2026

  Introduction to Graduate Algorithms

  Since the TAs don't have PhDs; and I don't think they should be TAing this course. I expect Professor to judge his TAs by giving them similar exam and see if they are qualified. So next time, TAs think twice about Honorlock.

  Rating: 1 / 5Difficulty: 5 / 5Workload: 37 hours / week

• ceqRuFTTtCeD9YNhW1KbbQ==August 3, 2026summer 2026

  Database System Implementation

  From a non-CS background, I didn't have any previous knowledge of OS, database systems, or other CS stuff. I checked the previous reviews and expected 15 hrs per week but ended up with 25 hours per week.

  The best part about the course is the assignments. There are 5 assignments. For assignments 1, 2, and 5, I took about 8–10 hrs each. For assignments 3 and 4, I took about 15–20 hours each. I also spent an extra 30 hrs learning C++ at the beginning.

  The lectures and other parts of the course are of low quality. The lectures are brief and don't cover much information. There is a requirement to read the textbook and papers every week, and that is the most time-consuming part. For the papers, there are research paper slides, and it is good to follow the points in them while reading. However, for the lectures, there are no clear instructions on what I need to focus on for the exam.

  The exam is tricky. I would say it is necessary to read the textbook and papers to pass the exam, as some of the questions are based on them. However, if you have a solid CS background, it may be a different story. About 90% of the information covered in the textbook is not tested, but I still have to read it because the remaining 10% is tested. I don't have enough knowledge to figure out which parts I should read and which parts I should skip. Some sections, for example Section 18, are assigned as reading, but they don't seem to be related to any part of the course.

  Overall, I learn a lot about OS, C++ and database system, but I don't enjoy the course since it does not provide enough guidance

  Rating: 2 / 5Difficulty: 4 / 5Workload: 24 hours / week

• ceqRuFTTtCeD9YNhW1KbbQ==August 3, 2026summer 2026

  Database Systems Concepts and Design

  From a non-CS background, I didn't have any previous knowledge of OS, database systems, or other CS stuff. I checked the previous reviews and expected 15 hrs per week but ended up with 25 hours per week.

  The best part about the course is the assignments. There are 5 assignments. For assignments 1, 2, and 5, I took about 8–10 hrs each. For assignments 3 and 4, I took about 15–20 hours each. I also spent an extra 30 hrs learning C++ at the beginning.

  The lectures and other parts of the course are of low quality. The lectures are brief and don't cover much information. There is a requirement to read the textbook and papers every week, and that is the most time-consuming part. For the papers, there are research paper slides, and it is good to follow the points in them while reading. However, for the lectures, there are no clear instructions on what I need to focus on for the exam.

  The exam is tricky. I would say it is necessary to read the textbook and papers to pass the exam, as some of the questions are based on them. However, if you have a solid CS background, it may be a different story. About 90% of the information covered in the textbook is not tested, but I still have to read it because the remaining 10% is tested. I don't have enough knowledge to figure out which parts I should read and which parts I should skip. Some sections, for example Section 18, are assigned as reading, but they don't seem to be related to any part of the course.

  Overall, I learn a lot about OS, C++ and database system, but I don't enjoy the course since it does not provide enough guidance.

  Rating: 2 / 5Difficulty: 4 / 5Workload: 24 hours / week

• Flq5Ybni4B0gY/9Ddy8jjQ==August 3, 2026summer 2026

  Advanced Topics in Software Analysis and Testing

  I took this course in the summer and it was a really manageable course. I roughly spent 10 hours per week. Some assignments were much simpler than others and didn't take too long. There was one exam in the middle of the semester which was okay in difficulty. Mostly it was along the same lines as the quizzes and materials already provided. The TAs were super prompt in grading and you get feedback very quickly.

  As for the course material itself, I didn't find it super engaging. I took this course as I thought it might be a good introduction to LLVM for the compilers course and that it was. I did gain some introductory understanding of LLVM and how to write some LLVM code. But the software analysis testing material itself was something that didn't interest me a lot. I found some parts interesting like the fuzzing lab and even the delta debugging part but other things felt a bit more theoretical.

  I recommend taking the course if you want lighter workloads (such as maybe for summer semesters) or want to gain some introductory knowledge about LLVM.

  Rating: 4 / 5Difficulty: 2 / 5Workload: 10 hours / week

• C5z0p8fJYg9Yks6o4/WdyA==August 3, 2026summer 2026

  Artificial Intelligence Techniques for Robotics

  This course has some well-made and fun projects. The most important thing about these is to start them early and make sure to check out the project resources for how to approach them. The TAs and professor also have frequent office hours to help you as well as project walkthroughs which are both very helpful.

  There are also problem sets that help you understand the concepts. Solutions will be provided for them and you could check your answers. I recommend working through them once for your own sake.

  The exams are also not that bad provided you understand the concepts thoroughly. You get two chances for each exam which is very generous. It is a pain to set up your environment for the room scan so make sure you have a personal space you can use.

  Overall, the course is not very difficult and you will walk away with a good experience in learning the material.

  Rating: 5 / 5Difficulty: 2 / 5Workload: 10 hours / week

• SSSOP28ZCXJKKGVI/KdIhg==August 3, 2026summer 2026

  Database System Implementation

  It's a fine class. You're going to learn a lot more about C++, design patterns, etc. than you will about databases.

  No need to read the papers/books. Lectures are enough. Exercise sheets and exams are fine, though there are an excessive amount of questions formatted as "which answer below is INCORRECT" which is annoying.

  Exams are essentially the exercise sheets + the practice questions given on Ed, with a few extra. You'll do fine if you just watch lectures and do the projects.

  For the projects you are given a skeleton and you just fill in methods. It is tedious more than anything else. You are provided with local unit tests, local Docker, etc. and you can submit to Gradescope many times. In my semester we implemented:

  • Lab 1, CSV manipulation: https://buzzdb-docs.readthedocs.io/part1/lab1.html
  • Lab 2, Buffer manager: https://buzzdb-docs.readthedocs.io/part1/lab2.html
  • Lab 3, B+-tree https://buzzdb-docs.readthedocs.io/part1/lab3.html
  • Lab 4, R-tree: https://buzzdb-docs.readthedocs.io/part1/lab4.html
  • Lab 5, queries: https://buzzdb-docs.readthedocs.io/part1/lab5.html

  Overall you will have a decent idea of how a toy database works. It will get you to understand the general data structures, file structures, etc.

  I imagine the majority of the class is doing the labs with AI assistance.

  Overall the class was OK. Not very challenging and unfortunately felt more like a class on a hodgepodge of topics. If you are looking for a good challenge w/ high payoff then I would suggestion distributed systems, compilers, GPU hardware/software over this course.

  Rating: 3 / 5Difficulty: 3 / 5Workload: 4 hours / week

• zTvc2jFZzQJrgASemhVtgg==August 2, 2026summer 2026

  Computer Networks

  I come across computer networks a lot in my Job even though it is not directly related to the field. Because of this I wanted a deeper understanding of computer networks. This course provided exactly what I wanted. I now have a solid understanding of the OSI network model, Content distribution networks, Video and Audio transmissions, data vs control plane etc.

  I might be one of the few who actually felt like I benefited more from the lectures and theory on the class than the actual projects. There are summary videos at the beginning of each section that I found helpful, but they only have them for the first half of the course. The rest of the lectures are just reading but I did learn a lot from them. There are quizes at the end of each lecture and they are open book and simple.

  I believe the projects are a little niche for my liking and will probably never work on something similar again in my life. Although they did help to get a deeper understanding of some topics. Like for example they have you configure a firewall to drop or allow different packets depending on the source an destination. Projects are relatively easy compared to projects on other clases.

  They are now introducing Guided Tutorials. I personally didn't like these. For our semester we only had to answer a quiz after doing the tutorial but I believe in the future they are planning for you to submit code files and implementations for these tutorials. The tutorial was easy to follow but required set up which was where I had some difficulties that turned a 2-3 hour tutorial to a 10h one. I think they are adding these just to make the class more difficult.

  They have 2 exams and they gave us a set of questions and answers to prepare which made the exams easy. I think the provided Q/A are really helpful to enforce the understanding of the material. Even if they didn't provide the Q/A if you are paying attention to the lectures you could probably score high on the exams.

  They provided lots of optional additional reading to deepen your understanding of the material.

  Overall a good class. On the easy side but you will definitely learn about computer networks.

  Rating: 4 / 5Difficulty: 3 / 5Workload: 11 hours / week

• rBAAprTd7n4xR3KjrheEEg==August 1, 2026summer 2026

  Introduction to Information Security

  This course serves as a good intro to cybersecurity with hands on learning. I came into this course with experience in software engineering and the course reinforced some useful takeaways that helped me in my job.

  Project breakdown:

  • Man in the middle (9 hours): Good intro to CTFs and you will be using Wireshark. Not too hard as long as you follow the storyline.

  • ML (4 hours): Crash course on sklearn, pandas, numpy. Not much learning

  • Binary Exploitation (20 hours): Extremely rewarding project and you get exposure to x86 assembly and manipulating the stack pointer. These flags were very tricky but the resources provided are very good. The project TA was super helpful during office hours

  • Cryptography (10 hours): While RSA and Vigenere Cipher were concepts I've seen, I liked getting exposure to attacking RSA in certain cases. Good review of number theory but the interesting challenge was working with very large numbers where one digit off in the precision doesn't allow you to pass Gradescope

  • API Security (6 hours): I had prior experience with using Postman and Swagger, so this was pretty easy. Extremely doable

  • Web Security (12 hours): I do full stack in my day to day, so the Chrome Developer Tools didn't feel foreign to me. One of the flags is pretty difficult and you have to get creative. This project like binary exploitation but applied to the application layer rather than low level layer

  • Log4Shell (6 hours): No prior background in JNDI, but as long as you watch the provided YouTube video, the project is easy. Good exposure to vulnerability in a common framework but the project instructions were clearly AI written which signaled lack of effort

  • Database Security (20 hours): One of my favorite projects because it goes beyond the usual SQL injection. This project forces hands-on application of how you can triangulate various datasets to make conclusions as well as use SQL injection to bypass poor client side and server side sanitization. I found it very useful and reinforces the boring annual cybersecurity trainings you get in the industry

  • Malware Analysis (6 hours): Not the most interesting project. Phase II of the project involved writing simple scripts to decode a string to get a flag. I felt this doesn't help me understand what's really inside malware. Also, the JoeSandbox environment in the VM is very laggy and the TAs anyway gave full score to the whole class on Phase I. At least one curve for the whole class

  General gripes about the course:

  1. You have 9 assignments with 1 week of time to complete each. Since summer semester is short, you are in constant time crunch and need to learn on the fly with the provided resources

  2. The VM setup for Mac users is horrible. In some projects, the VM is extremely laggy which slows you down. UTM consumes nontrivial resources on an M4 macbook and only later in the course did a TA mention a cloud based setup. Future iterations of the course should use cloud based setup and I believe GT can provision the computing resources for this class

  3. The course was expected to give writeup after the projects are completed to reinforce the learning objectives, but only 2/9 projects came with the writeup. Guess the course didn't honor its own syllabus

  4. Limited Gradescope submissions on some projects added artificial difficulty.

  Rating: 4 / 5Difficulty: 3 / 5Workload: 12 hours / week

• RJnVGg81f1ifSy69onW2IA==August 1, 2026summer 2026

  Game Artificial Intelligence

  Well-designed course. There are 8 assignments with clear instructions and then a bunch of quizzes due at the end. Each assignment corresponds well to the lectures and clearly teaches a concept. Lectures are very long, it can be easier to look through slides or transcripts.

  Rating: 5 / 5Difficulty: 3 / 5Workload: 8 hours / week

• 4ID5lelLmpAK7+R5a58ySw==August 1, 2026summer 2026

  Database System Implementation

  This course covers an interesting and useful topic, and it is mostly effective in teaching the material. The course has two exams, two exercise sheets (which are basically just quizzes), and five programming assignments.

  The meat of the course is in the programming assignments. This was where I learned the most about how databases work under the hood, and the reasons for why certain design decisions are made. None of the assignments were too difficult, but I would still start them early, as understanding the requirements and boilerplate code can take a bit of time. By the end of the course, you'll have built a very rudimentary DBMS, which is rewarding.

  The lectures and exams I felt were more of a miss. They very strongly emphasize C++ syntax, which seems odd for a topic that should really be language agnostic. I enjoyed reading the textbook and papers, and I wish the lectures focused more on those topics rather than teaching introductory programming concepts.

  I do think this course is worth taking if you enjoy programming assignments and are less worried about the quality of exams/lectures. If you want to learn about database internals and go in with the mindset that this is a coding heavy class, you will likely enjoy it.

  Rating: 3 / 5Difficulty: 2 / 5Workload: 10 hours / week

• 4ID5lelLmpAK7+R5a58ySw==August 1, 2026summer 2026

  Introduction to Graduate Algorithms

  This was a rather difficult course that teaches the material well and evaluates students in a stressful but ultimately effective way. Your grade is broken down into 90% exams and 10% quizzes. To pass with a B, you need to average at least a 40/60 on the exams assuming full points on the quizzes. The heavy weighting of the exams certainly makes the stakes high, but I can understand why they have decided on this format.

  The exams are split into free response questions and multiple choice questions. I found that the homework reflected the style of reasoning expected for the FRQs. And the multiple choice questions are rather easy to intuit if you've done the homework.

  Expectations for the class are very clearly laid out, which I appreciated. For each exam, the TAs make multiple posts explaining how to structure your algorithms, common ways people lose points, and what assumptions you can and cannot make. It's a great deal of information to parse, but the exams and grading should have no surprises if you've done your due diligence.

  I personally loved this class and am glad I took it. If you treat it as an opportunity to delve into the heart of computation at a graduate level, you will be rewarded.

  Rating: 5 / 5Difficulty: 4 / 5Workload: 20 hours / week

• kNC4fH+DJPa4k7hkAIIzUA==August 1, 2026summer 2026

  Database System Implementation

  This course has potential to be up there with the more recognized courses, like GIOS, AOS, HPCA, etc. but falls pretty short. The lectures are not and will not be relevant to exams. They provide a high-level overview of the topics and you're better off skipping them. You can expect to gain more relevant knowledge to the programming assignments through the designated readings, which to give credit, was pretty interesting albeit dense. The exams are gimmicky: I don't like the wording of the questions and I feel like it's meant to trip you up more than it does testing your knowledge on database systems.

  The best things about this course are its programming assignments, TAs, and the passion the professor has for this topic. The programming assignments are all in C++, but does a good job of giving you the proper harness to run and test. Great developer experience. The TA is very friendly and nice, fast responses too.

  I got a C in this course, so take my opinion and review from a student that didn't expect to give much effort in this course.

  Rating: 2 / 5Difficulty: 4 / 5Workload: 12 hours / week

• 0qNdCyeDuSTnuVthrbLt9g==July 30, 2026spring 2026

  Introduction to Graduate Algorithms

  This course is not about teaching but about Georgia Tech making it difficult for students to graduate. Why? This is how they make their money. You can only graduate if you pass with B or better. Your grades are only depended on quiz/exams. The exams are made difficult and I am 100% positive that the Head T.A cannot even pass it himself given his credentials. I ask the Professor to test his TAs on these exams; so next time when TAs are making rules on how to protract students; they think twice.

  Rating: 1 / 5Difficulty: 5 / 5Workload: 45 hours / week

• 0qNdCyeDuSTnuVthrbLt9g==July 30, 2026summer 2026

  Introduction to Graduate Algorithms

  AVOID THIS CLASS UNLESS YOU WANT TO REPEAT THIS COURSE.

  Rating: 1 / 5Difficulty: 5 / 5Workload: 37 hours / week

• 0qNdCyeDuSTnuVthrbLt9g==July 30, 2026summer 2026

  Introduction to Graduate Algorithms

  There are 3 exams and 8 Quiz for this course. Homeworks are optional and don't count towards your grade. My biggest concern with the course is the emphasis on exam proctoring and strict procedural requirements. A significant amount of attention seems to be placed on preventing and detecting cheating, rather than creating an environment that encourages students to learn and understand the material. I also found the role of the TAs frustrating. In my experience, the TAs were heavily involved in administering and evaluating course matters and having zero academic credentials to judge anyone.

  The lecture videos are also old and do not necessarily reflect the current instructor's teaching approach. Many of the underlying algorithm lectures are also available for free on YouTube, which raises the question of what additional instructional value the course is providing. A lot of students had to repeat the course; and the question is why is this the case? if your students are repeating the class, it shows you are not a good instructor and no one is learning.

  I know its hard to avoid this course because it is required for graduating but DO WHATEVER IT TAKES TO HAVE THE DEAN REMOVE THIS COURSE OR REMOVE THE INSTRUCTOR/TAs AND PLACE BETTER PEOPLE.

  Rating: 1 / 5Difficulty: 1 / 5Workload: 40 hours / week

• /gsnycu00Mi5btTTN3MgyQ==July 29, 2026spring 2026

  Cyber Physical Design and Analysis

  I learned a lot in this course. Many of the concepts were new to me. It's more of a survey course where you learn about the wide variety of topics in cyber physical design and analysis, but you don't jump too deeply into any one topic.

  Projects 1 and 2 and homeworks 3 and 4 are a lot of fun, and are relevant to the course material/lectures. Several of the reviews mention that the lectures are not necessary for completing the assignments, but if you don't watch the lectures then the assignments will be more difficult. The lectures provide techniques that you can utilize that will make your life SO much easier in the assignments (Finite State Machine for projects 1 and 2 for example). I recommend watching the lectures for everything other than AADL. It is easy to miss a lot of points on the projects and homework if you don't pay attention, and don't watch the lectures.

  Project 3 and homework 5 are not very enjoyable. However, when I took this class the TA's and professors had just re-worded every assignment to make it more clear what is required. I believe this helped a lot as I got full points for homework 5 and Project 3 pt 2, and nearly full points for Project 3 pt 1. Even with the re-wording, I don't think that these projects should be in the course. I think there are probably better options as to projects and homework that can drive the point home better for the concepts they're trying to convey. Working with AADL was an immense pain which made it a total slog.

  This is a harder-than-average (but not by too much) OMSCS course in my opinion. I ended with a high A in the class. It's probably not too difficult to pass, but to get an A you will have to put in effort. I did learn a lot, so it was definitely worthwhile to take.

  Also, all the reviews mentioning that this course has nothing to do with cybersecurity - yeah duhh. The word "cyber" does not automatically imply "security". There are many different topics in cyber. No one should be surprised by this when registering for this course. You can tell who doesn't pay attention to what they register for which might be good context on how to interpret those reviews.

  Rating: 3 / 5Difficulty: 3 / 5Workload: 12 hours / week

• /gsnycu00Mi5btTTN3MgyQ==July 27, 2026summer 2026

  Video Game Design and Programming

  This class is very fun and I highly recommend it. I went into the class just because it was highly rated and it was supposed to be easy, I had no interest in video games or in making one. However, learning the various aspects of making a game, as well as actually making one for the group project was LOADS of fun. You get what you put into it. And you definitely will put in a lot of time on the group project. I spent a ton of time on the group project because my group was slacking a lot. However, even though I had to pick up the slack, I still thoroughly enjoyed making the game and working with unity. The concepts are not difficult, but you will probably spend (or should spend) more time than you think you'll need to spend on this class purely because of the group project.

  Rating: 5 / 5Difficulty: 3 / 5Workload: 20 hours / week

• T52KwjMZVFf7TnQWkbHiTg==July 27, 2026spring 2026

  Graduate Introduction to Operating Systems

  (I actually took this class a while ago, but apparently you can't go back that far)

  I took this as one of my first two courses in the program along with Introduction to Graduate Algorithms. I came into the class having already taken an operating systems course during my undergraduate degree, but still found the material to be rather eye-opening and made me interested in taking more systems courses.

  I received 100% on all three of the projects. They were challenging, and had a steep learning curve, but were fair. The exams felt fair as well; nothing was too surprising if you had kept up with the lectures and understood the material.

  Overall, I learned a lot about IPC, threads, processes, memory management, and networking. The projects can take some time, especially if you don't have much experience with C, but I found them to be extremely useful for reinforcing the material.

  Overall, I highly recommend taking this class at some point during your time in OMSCS!

  Rating: 5 / 5Difficulty: 3 / 5Workload: 17 hours / week

• lEeozhOiSvhx9PyqXN1IWQ==July 27, 2026summer 2026

  Artificial Intelligence Techniques for Robotics

  I liked this course. I hadn't taken a course that touched on any of these subjects: localization, mapping, navigation, control systems.

  So to me this was a really good class. However, I think it's not a very in-depth class. Which is why I rated it 4/5. If you had already been exposed to any of those concepts than this class won't add to your understanding. They do offer some extra credit opportunities that would give you some more in-depth understanding. I didn't do them because I was a little burnt out (I've taken GIOS, HPC, AOS, HPCA, DC, SDCC). After DC and SDCC I was just needing a bit of a break you know?

  I spent ~7 hours a week during the summer. For a normal semester you can probably get sub 5 pretty easily. It's a nice easy class. I will say it seemed like a lot of other students struggled more than me. I'm not sure why. The TAs give really good instructions and walkthroughs for the projects, but more importantly the HWs are like 90% of what the project is. For projects 5 and 6 ~85% of my code was copied from the HW. I think people didn't step back enough and think about how the project was different from the HW and what to keep vs what to update between the two which is why they struggled. 30-45 minutes of planning saves you hours.

  100% of my code was hand written instead of using some AI assistance, but that gives you an idea of the 7 hours a week part of it. If anything I think AI would have over engineered most of the projects and probably wasted time. Maybe that's another reason other students struggled.

  I think the difficulty is is 2/5. There are some parts that can be challenging (smoothing project). But I rated it as 1/5 because to balance out some of the other crazy high ratings for this course.

  Overall I'd recommend it immensely for anyone who hasn't been exposed to the subjects or needs a nice easy class in-between some of the monster time sinks in the program.

  Rating: 4 / 5Difficulty: 1 / 5Workload: 7 hours / week

• K/pqMgtPVqlyo2QMKauVbA==July 27, 2026summer 2026

  Machine Learning for Trading

  I did not enjoy this course at all. I came in excited to learn about ML and Finance, but I left feeling completely overwhelmed.

  This course is ruined by the excessive workload, the stubborn attitude of the TAs, and the way the exams are designed. The TAs refuse to answer straightforward questions and instead point you to discussion posts or course materials, even when you're looking for clarification. The exams focus on whether you remember details from the lectures rather than whether you actually understand the concepts. The exams also use finance technical jargon all the time, making it feel like it's all about identifying the tricky wording (and less about understanding the concepts).

  The assignments were also much heavier than they needed to be because of the overwhelming number of requirements. Instead of reinforcing the concepts being taught, they emphasized implementation complexity and busywork. Why does every project description need to be 10 pages long? Every assignment in this class felt unnecessarily time-consuming, even when the underlying technical concepts were relatively simple.

  Rating: 2 / 5Difficulty: 4 / 5Workload: 20 hours / week

• BrHCAwk8rHVMLMavnc4MiA==July 25, 2026summer 2026

  Data Analytics in Business

  This course feels a bit rudderless to me. I took it as the 5th course in the OMSA program, and I didn't feel like any of the material was particularly new or interesting. Along with this, this course feels like it's lacking some kind of overall purpose. The homeworks were okay, but the demo videos basically do every assignment for you. I felt no real incentive to dive into learning anything deeper. Overall, I would say this course is far too easy for a master's program, and overall, I'm a bit disappointed. I think the lecture is a great base for the course, but there needs to be more to it.

  Rating: 2 / 5Difficulty: 1 / 5Workload: 3 hours / week