## Instructor:

## Semester:

- 2012 Autumn/Monsoon (Aug - Dec)

Course Objectives: Learn the basics of estimation theory, and ma-

chine learning algorithms. The key topics we will study are as follows.

Supervised Learning: Linear Regression, Linear classication, Model

Selection and Inference, Support Vector Machines, Graphical Models

and Message Passing, Principal Component Analysis, High dimen-

sional statistical learning.

Unsupervised Learning: Clustering, Cluster Analysis, Multidimen-

sional Scaling.

Compressed Sensing: Sparse Signal Processing and Inference.

Textbook: We will follow "The Elements of Statistical Learning"

Hastie, Tibshirani and Friedman, and Pattern Recognition and Ma-

chine Learning by Christopher Bishop.

Additional References: Topics in Matrix Analysis: Horn and John-

son

Course Project:

A team of two students will be assigned a single semester-long

research project. The project's purpose is help you build exper-

tise in an area related to machine learning to the point of being

able to contribute original research. The area of your project

will be chosen by you with instructor's guidance. The project

has four main goals: 1) to build your background knowledge by

having you read 3-5 papers in your chosen area 2) to learn writ-

ing skills/techniques by writing a summary of the most of the

read papers 3) to learn speaking skills/techniques by presenting your area to the class 4) to push you to think about original con-

tributions to the area by working on extensions of the material

you learn. Last being the most important and carries the

maximum weight.

- Course Project Meeting: Each team will meet with the instuctor once every week for an hour.

- Towards the end of the course each member of the team will give a presentation to the class on your selected area.

- Each team will write a summary of the read papers using LaTeX.

Grading: There will be 1 mid-term test and a nal/ just one nal

with equal weightage of 15%=30%. The course project will be worth

40% of your grade. The rest 30% will comprise of homeworks, and

class performance.

Auditing Policy: Students planning to audit the course will be re-

quired to turn in all the homeworks, and the course project. Writing

exams is, however, optional.