Basic Information
- Course Code: EE 782
- Course Name: Advanced Topics in Machine Learning
- Course Offered In: 2023-‘24
- Semester Season: Autumn
- Instructors: Prof. Amit Sethi
- Prerequisites: Any introductory ML course
- Difficulty (1 being easy and 5 being tough): 2
Course Content
The course covered a wide variety of ML models and techniques, including various CNN models, RNNs, LSTMs, generative models, Transformers, basic NLP, robust loss functions, graph neural networks, sem-supervised learning, and uncertainty estimation. Due to this massive breadth of concepts, each of the topics was covered in a single lecture or two.
Feedback on Lectures
The lectures were decently well-paced and provided a good overview of each of the topics covered. The lectures did not go much into the mathematical concepts behind these models and mostly dealt with the application side.
Feedback on Evaluations
Grading Scheme was approximately: 28% Assignments, 8% Quiz, 18% Project, 18% Midsem, 28% Endsem. There were two assignments, both of which were long and laborious, but the students were provided with enough time to do them. There were 5 minutes long weekly quizzes, which were just a way of taking attendance in class. Both the exams were open book and were on the easier side, with a large number of students scoring close to full.
The grading seemed more than generous as compared to the other courses, but was mostly on the expected lines, as the instructor is known for his policy of absolute grading, and a large number of students this year had high marks.
Study Material and Resources
The instructor shared numerous recently published deep-learning papers to complement the contents of the lecture slides.
Follow-up Courses
CS 726 (Adv ML), students may also find themselves interested in other deep learning-based courses like Intro to NLP, CV and ASR after doing this course.
Final Takeaway
This course is particularly useful for students who want to work in the field of ML applications. The breadth of concepts covered in the course will be beneficial to anyone sitting for ML interviews, and overall this course does provide an introduction to various topics that are covered in much more detail in their specialized courses.