Foundations of Machine Learning

Course#: CSCI-GA.2566-001

Instructor: Mehryar Mohri

Graders/TAs: Vitaly Kuznetsov, Andres Munoz Medina, Scott Yang.

Mailing List

Course Description

This course introduces the fundamental concepts and methods of machine learning, including the description and analysis of several modern algorithms, their theoretical basis, and the illustration of their applications. Many of the algorithms described have been successfully used in text and speech processing, bioinformatics, and other areas in real-world products and services. The main topics covered are:

- Probability tools, concentration inequalities
- PAC model
- Rademacher complexity, growth function, VC-dimension
- Perceptron, Winnow
- Support vector machines (SVMs)
- Kernel methods
- Decision trees
- Boosting
- Density estimation, maximum entropy models
- Logistic regression
- Regression problems and algorithms
- Ranking problems and algorithms
- Halving algorithm, weighted majority algorithm, mistake bounds
- Learning automata and transducers
- Reinforcement learning, Markov decision processes (MDPs)

Location and Time

Warren Weaver Hall Room 109,

251 Mercer Street.

Mondays 5:10 PM - 7:00 PM.

Prerequisite

Familiarity with basics in linear algebra, probability, and analysis of algorithms.

Projects and Assignments

There will be 3 to 4 assignments and a project. The final grade is essentially the average of the assignment and project grades. The standard high level of integrity is expected from all students, as with all CS courses.

Lectures

- Lecture 00: Introduction to convex optimization.
- Lecture 01: Introduction to machine learning, basic definitions, probability tools.
- Lecture 02: PAC model, guarantees for learning with finite hypothesis sets.
- Lecture 03: Rademacher complexity, growth function, VC-dimension, learning bounds for infinite hypothesis sets.
- Lecture 04: Support vector machines (SVMs), margin bounds.
- Lecture 05: Kernel methods.
- Lecture 06: Boosting.
- Lecture 07: Density estimation, Maxent models, multinomial logistic regression.
- Lecture 08: On-line learning.
- Lecture 09: Ranking.
- Lecture 10: Multi-class classification.
- Lecture 11: Regression.
- Lecture 12: Reinforcement learning.
- Lecture 13: Learning languages.

Textbook

The following is the required textbook for the class. It covers all the material presented (and a lot more):

- Mehryar Mohri, Afshin Rostamizadeh, and Ameet Talwalkar.
*Foundations of Machine Learning*. MIT Press, 2012.

Technical Papers

An extensive list of recommended papers for further reading is provided in the lecture slides.

Homework

Previous years