Your feedback has been sent to our team.
8 Ratings
Hours/Week
No grades found
— Students
Sections 1
Professor Nguyen is honestly quite a poor lecturer. His slides are sparse and explanations are quite inadequate but if you have sufficient ML background, this isn't a huge deal. The codeathons and assignments are pretty easy, only taking maybe an hour or two to complete. The exams (just a midterm and the final) are a bit difficult but don't count for that much.
This was a very good class to learn various ML models and code them in Python. Rich goes into a lot of detail during his lectures and a lot of the math went straight over my head. I found the lectures kind of boring, but the coding assignments were rewarding. You learn a lot about different ML models like Random Forest, Decision trees, etc. in the first part and neural networks in the second part of the course.
Grading is 40% programming assignments (4), 15% codeathons (3), 30% exams (midterm and final), and 15% the group project. The schedule is a little annoying in that there'll be no work for a couple of weeks and then a bunch of assignments in a row. The programming assignments aren't super difficult but will probably require you to go to OH a couple of times for help. The codeathons are more challenging and also take longer to run, so I would definitely start those as early as possible (ESPECIALLY codeathon 2 since it's convolutional neural nets). The coding is a mix of implementing algorithms from scratch and using built-in libraries like ScikitLearn and Keras. The final project is also interesting since we implement the models we learned on a dataset of our choosing but tbh only the first codeathon is relevant. Do as much extra credit as you can because then you won't stress as much about the assignments.
The midterm was a mix of application and conceptual questions, and it was online and open-note. For the conceptual questions, you can use your notes, but the application questions were unnecessarily time-consuming (like he made us calculate the information gain for 4 different variables when it really only takes one or two to check if we understand how it works.) Other than being a little time-consuming, I didn't think the midterm was horrible (ended up with over a hundred on it since there was a bonus question).
#tCFfall2021
Nguyen is a great guy and a solid lecturer. Take the class if you're interested in ML. He goes into a lot of detail on how each algorithm works conceptually. He also elaborates a lot on the mathematical formulas, which I didn't really care for. This class wasn't too interesting to me personally. It was a good survey at ML though, and I learned a lot theoretically. Don't ask me to implement a regessor on a dataset, though.
The coding was so under the hood, which is unnecessary imo and I wish we coded in a more practical manner than implementing each major algorithm from scratch when there's proven ones that work better out there, but I guess it did help me understand the ML algorithms better. The exams were challenging, as he asked math questions and fine details. The second codeathon was also incredibly stressful, but I got lucky and some classmate shared a tip with some of us to get insanely high performance at OH. The final project is cool bc you can apply all the ML functions to a dataset you've cleaned and see if you can find any insights. The video (though we went online) was fun too.
DO THE EXTRA CREDIT!!! Seriously, it will save your grade. Trust. Me. That's the only way I ended up with an A.
[TAKEN ONLINE]
If you are interested in applying Machine Learning principles practically in industry, this is the class for you. This class is primarily focused on giving you a somewhat barebones foundation for the theoretical aspects of machine learning while introducing you to many commonly used open-source libraries such as sklearn, keras, and Tensorflow. Professor Nguyen's lectures consist primarily of him playing prerecorded videos of him explaining topics while answering questions in the Zoom chat, which can be somewhat tedious and difficult to pay attention to, but everything is recorded and posted so it isn't too bad.
The course covers a ton of material but doesn't go into a lot of depth. There are a lot of equations thrown at you without a lot of explanations. Assignments could be a bit frustrating sometimes, and it was really hard to get TA help. It would definitely have helped if this class was separated into two classes so that he could have enough time to cover everything with proper detail. Still, though, I appreciate how much the professor pours his heart and soul into this class, and he's definitely very accomodating. The lectures are good and flow really well. Well worth your time.
This course was a very well-rounded survey of machine learning and deep learning concepts. Professor Nguyen is very accomodating and constantly checked in with us to see if we needed deadlines extended. The class project was also a great applied learning experience and has helped me land a lot of interviews!
Get us started by writing a question!
It looks like you've already submitted a answer for this question! If you'd like, you may edit your original response.
No course sections viewed yet.