Machine Learning
ES 654 · IIT Gandhinagar · Semester II, 2019–20
ES 654 introduces machine learning from decision trees and ensembles through linear regression, optimisation and regularisation, Bayesian learning, logistic regression, hidden Markov models, entropy and Gaussian processes, to neural networks, support vector machines, naive Bayes, nearest neighbours, unsupervised learning and reinforcement learning. Every class was recorded.
- Instructor
- Prof. Nipun Batra
- nipun.batra@iitgn.ac.in
- Classes
- Tuesday and Friday, 3:30–5:00 pm, 1/101
- Office hours
- Monday, 12:00–1:00 pm
- Teaching assistants
- 4 TAs
- Prerequisites
- Good experience in Python programming, probability and linear algebra
Lectures in this offering ran from January to July 2020. Its slides and lecture recordings remain available.
- Lectures
- 32
- Slides
- 36 slide decks
- Recordings
- YouTube playlist
- Grading
- Project, homework, quizzes and more
Lectures
- 1–5Foundations and decision trees5 lectures
- 6–13Linear regression and optimisation8 lectures
- 14–20Bayesian and probabilistic models7 lectures
- 21–28Neural networks and support vector machines8 lectures
- 29–32Naive Bayes, nearest neighbours and further topics4 lectures
Questions
Are the lectures recorded?
Yes. Every class has a recording on YouTube, linked from its row on the Lectures page; the full set is in one playlist.
Is there a project?
Yes: a group project in teams of up to four, worth 40%, assessed through a proposal, two presentations, a three-minute pitch, a demo and poster, and a final report. Ideas: from the instructors and from a Stanford ML course.
How do paper presentations work?
Three students take each paper: one summarises it, one argues for it, and one acts as devil's advocate, convincing the jury that the paper has flaws. See the paper list and the slides template.
Can I submit homework late?
For 50% credit: up to one day late for the first assignment and up to two days for the others. A subset of assignments has a viva.
How do I prepare before the course?
Refresh the prerequisites with the first four chapters of the Python Data Science Handbook, some material on linear algebra and Khan Academy's statistics and probability course.
References
Reference textbooks:
- Gareth James, Daniela Witten, Trevor Hastie and Robert Tibshirani, An Introduction to Statistical Learning with Applications in R (online)
- Christopher M. Bishop, Pattern Recognition and Machine Learning, Springer, 2006 (free PDF)
- Jerome Friedman, Trevor Hastie and Robert Tibshirani, The Elements of Statistical Learning, Springer, 2001 (free PDF)
- Richard O. Duda, Peter E. Hart and David G. Stork, Pattern Classification, Wiley, 2012
- Tom M. Mitchell, Machine Learning, McGraw Hill, 1997
- Kevin P. Murphy, Machine Learning: A Probabilistic Perspective, MIT Press
- Ian Goodfellow, Yoshua Bengio and Aaron Courville, Deep Learning, MIT Press, 2016 (online)
Other ML courses:
- NPTEL course by Balaraman Ravindran
- CMU course by Tom Mitchell and Maria-Florina Balcan
- Coursera course by Andrew Ng
- fast.ai course on ML and Practical deep learning for coders
- Course by Alex Ihler, UCI
Teaching assistants
- Indradeepindra.mastan@iitgn.ac.in
- Sudhakarsudhakar.kumawat@iitgn.ac.in
- Supratimsupratim.shit@iitgn.ac.in
- Shubhamshubam.singh@iitgn.ac.in