Instructor

Prof. Amit Sethi

Semester

Spring ‘20

Course Difficulty

I would rate this course moderately difficult for those who have no previous exposure to ML. It should be relatively easy for the students who have done online ML courses. This course content, though not tough, is very vast. A good amount of time is required to go through all the lectures. In addition, you have to complete graded tutorials and lengthy programming assignments. Project has significant weightage too. Loadwise, the course is very heavy.

Time Commitment Required

8-9 hrs/week

Grading Policy and Statistics

The grading is very reasonable. Absolute Grading with 90+ AA, 80+ AB and so on. Due to significant bonus marking, many students received AA/AB. If you put decent efforts, then the grade definitely won’t disappoint you.

Attendance Policy

I took this course in the online semester. Attendence was not compulsory. Most of the classes were for discussing doubts pertaining to the uploaded lectures.

Pre-requisites

There is no hard prerequisite to this course. That being said, having some exposure to ML such as coursera courses would give you a definitive advantage. Python programming is required in all assignments and project.

Evaluation Scheme

The course has tutorials, assignments, midsem, endsem, project.
Midsem ~ 20%
Endsem ~ 30%
Project ~ 25%
Assignments ~25%
Tutorials ~ 15%

Course total is out of ~ 115

Topics Covered in the Course

This course covers all the basic ML concepts including topics such as Linear and Logistic Regression, SVM, Kernels, Decision Trees, Neural Networks, Clustering, PCA, Expectation Minimization Algorithm.

Teaching Style

The prof uploaded prerecorded lectures (70+ videos averaging ~20 mins per video). Regular classes were held in the given slot. They were mainly doubt solving sessions which need not be attended if you dont have serious doubts. Personally, I found the lecture videos to be perfectly adequate for conceptual understanding. Professor’s teaching is very thorough and you won’t need to refer to any other source.

Tutorials/Assignments/Projects

We had 6 tutorials which were quite mathematical. There were 3 assignments which took about a week each to complete. The project is accompanied with an in-depth viva, so be prepared.

Feedback on Exams

The exams are moderately difficult and thought provoking. No rote learning is required. Professor puts a lot of thought into the questions. Due to bonus marking, getting decent scores should be easy.

Motivation for taking this course

This course is meant to introduce the students to the basics of ML. It has both theoretical and practical aspects. After completing this course, you are eligible to take many advanced DL / AI courses.

Course Highlights

Very informative, Lengthy, Covers a lot of ground, easy AA with consistent efforts. Great course overall.

Course Importance

This course is very important if you want a career in AI.

How strongly would I recommend this course?

If you want to pursue a career in AI/DS, this course is a must. I also believe basic ML awareness will be required in most technical roles in the upcoming years. Hence, I would strongly recommend everyone to take EE769/DS303 during their time at IITB.

When to take this course?

I took this course in my 6th Semester. If you have made up your mind about a career in data science early on, then 4th semester would be the best time to take this course. This allows you to cover most of the advanced courses in the following years.

Going Forward

Advanced ML courses/ DL courses. Having completed this basic ML course, one can approach professors for research projects.

References Used

-

Other Remarks

-

Interesting relevant links

-

Review By: Vikrant Rangnekar