He was a good professor, always ready to help and responsive to emails. The first couple homeworks were kinda hard, but he was available in office hours. Attendance is mandatory for lecture and discussion, as they randomly give quizzes to check. The quizzes are easy and marked for completion. This professor emphasizes the math behind ML stuff, so be ready for that. It's not a huge deal if you don't come from a math background, but the beginning stuff might be confusing. Prof and TA are there to help though!

I would recommend this class! Grade was 40% homework, 40% group project, and 20% quiz.

The quizzes would be held in class or in discussion and they were super short like 1 question on a basic topic like matrix multiplication or drawing a neural network. Half of the quiz grade was on attendance and half on correctness. He dropped 6 of the quizzes and they were really easy, they were just to check attendance.

There were 5 homeworks on data visualization, gradient descent/SGD, linear algebra/PCA stuff, neural networks, and clustering. The first two were pretty hard but the rest were pretty easy.

The group project didn't have strict requirements, you could do it on pretty much any data analysis/science topic but most people chose to train models for tasks like image classification etc.

Grading wasn't too strict for the most part, most people got As on the project. The professor is also super helpful and if you go to office hours he will help you through the homework/project. I didn't find discussion too helpful but I went in case there was a quiz.

Topics taught were Plotly (data visualization), neural networks (Tensorflow and PyTorch), clustering (scikit learn), linear algebra stuff and doing PCA by hand, network data science, and some other topics I didn't pay attention to because there were no more homeworks lmao.