EE769 – INTRODUCTION TO MACHINE LEARNING
Course: EE769 – Introduction to Machine Learning
Semester: Spring Semester, 2018
Professor: Amit Sethi
Course Content: Statistical Learning Theorey, PAC learning, different loss functions and properties, training, validation, and testing. VC dimension, Bias-variance tradeoff. Supervised learning algorithms: Linear regressions, Perceptron logistic regression. SVM, support vector regression, decision trees Kernelization of linear problems: increase in dimensionality through simple kernels, kernel definition and Mercer’s theorem, Kernelized SVM and SVR. Neural Networks: introduction, weights biases etc, back propogation. Other applications of kernels. Random forests. Unsupervised learning algorithms: Clustering criteria, K-means, DB-scan, EM-algorithm. Kernel-PCA, introduction to generative and probabilistic graphical models. Bayesian networks.
Evaluation Criteria: Couple of quizzes, 1 compulsory assignment and another assignment + term paper (with various topic options) against one project. Regular midsem and endsem.
Feedback on Lectures: Professor roughly followed the slides and a couple of reference books mentioned for the course. As with most of the other courses, it will be a heavy burden if lectures are missed. The lectures involve knowledge of probability and course content is slightly heavy on mathematics.
Feedback on Assignments, Tutorials and exams: Quizzes and regular midsem and endsem are centered around the content taught during class. The assignments were based on two major applications, supervised and unsupervised learning. There was considerable freedom in choosing the project topics. Also, the term paper topic choices were very interesting. Apart from core ML related topics, the professor also offered the freedom to write about other aspects of ML like social or economical.
Difficulty Level: Slightly heavy
Grading: Please refer ASC. In general, moderate.
Textbooks and References: 1. Understanding Machine learning from Theorey to Algorithms, Shai Shalev-Shwartz and Shai Ben-David 2. Pattern Recognition and Machine Learning, by Christopher Bishop