Introduction to Machine Learning

Course outline

Taught by Prof. Balaraman Ravindran of IIT Madras, this NPTEL video series runs to 71 lectures. It starts with the basic paradigms of learning (supervised, unsupervised and reinforcement) and then works through statistical decision theory for regression and classification.

The middle of the course

A large block covers the toolkit around regression: bias-variance trade-offs, linear and multivariate regression, subset selection, shrinkage methods, principal components regression and partial least squares. The series closes on linear classification.

Who it suits

Students and engineers who want a grounding in classical machine learning, rather than a tool-focused tutorial, will get the most from it. Because it is a lecture series of 71 parts, pacing it over several weeks is a sensible plan.