Machine Learning

Computer Science CSCI-B555

Spring 2015

 

How to prepare for the class: see here

Syllabus

Midterm Exam: Week 9, March 10th, in class.

Final exam is scheduled for Thursday May 7th, 2:45pm-4:45pm, in LH008. Official IU exam schedule.

Office Hours AFTER the final exam will be on Friday May 8th, 4pm-5:30pm, in LH301F.

 

 

Instructor's lecture notes: 1, 2, 3, 4, 5, 5 (updated 04/03) 6, 7, 8, 9, 10, 11.

Instructor's slides: 1, 2, 3, 4 (from a different course), 5 (updated on 5/1).

Students' lecture notes from 2010: 1, please be careful, some have errors.

 


Week 1: January 12-16, 2015

 

Topics

Class introduction

Review of probability theory

 

Reading material

Textbook: Introduction (Chapter 1)

 


Week 2: January 19-23, 2015

 

Topics

Random variables

Introduction to parameter estimation

 

Reading material

Textbook: Introduction (Chapter 1)

 

(Homework Assignment #1)


Week 3: January 26-30, 2015

 

Topics

Parameter estimation

    Maximum a posteriori (MAP)

    Maximum likelihood (ML)

    Bayesian principles

 


Week 4: February 2-6, 2015

 

Topics

Expectation-maximization (EM) algorithm

 

Reading material

Textbook: Mixture models and EM (Chapter 9)

 

Code from class

Matlab code for the EM algorithm.

 


Week 5: February 9-13, 2015

 

Topics

Basics of classification and regression

Ordinary least-squares (OLS) regression

 

Reading material

Textbook: Introduction (Chapter 1)

Textbook: Linear models for regression (Chapter 3)

 


Week 6: February 16-20, 2015

 

Topics

Algebraic view of linear regression

Non-linear regression using OLS regression

Regularization

 

Reading material

Textbook: Linear models for regression (Chapter 3)

 

Code from class

Matlab code for the linear regression.

 

(Homework Assignment #2)


Week 7: February 23-27, 2015

 

Topics

Bayesian linear regression

Logistic regression

 

Reading material

Textbook: Linear models for classification (Chapter 4)

 


Week 8: March 2-6, 2015

 

Topics

Perceptron

The Pocket algorithm

 

Reading material

Textbook: Linear models for classification (Chapter 4)

The Pocket Algorithm can be found from here

 

Code from class

Matlab code for the logistic regression.

Matlab code for the pocket algorithm.

 


Week 9: March 9-13, 2015

 

Midterm Exam, Tuesday, in class.

 

(Class Project)


Week 10: March 16-20, 2015

 

Spring Break!

 


Week 11: March 23-27, 2015

 

Topics

Naive Bayes classifiers

 

Reading material

Textbook: Linear models for classification (Chapter 4)

Tom Mitchel's preliminary book chapter on Naive Bayes classifiers is here.

 


Week 12: March 30-April 3, 2015

 

Topics

Data-preprocessing

Performance estimation

 

Reading material

Jiawei Han's slides are here.

Tom Mitchel's book chapter on accuracy estimation, posted in Oncourse.

 

(Homework Assignment #3) (HW3 data)


Week 13: April 6-April 10, 2015

 

Topics

Classification and regression trees

 

Reading material

Tom Mitchel's book chapter on decision trees, posted in Oncourse.

 


Week 14: April 13-April 17, 2015

 

Topics

Classification and regression trees

Neural networks

 

Reading material

Textbook: Neural networks (Chapter 5)

Tom Mitchel's book chapter on decision trees, posted in Oncourse.

 


Week 15: April 20-April 24, 2015

 

Topics

Neural networks

 

Reading material

Textbook: Neural networks (Chapter 5)

RPROP paper available here.

Papers by Breiman and Freund & Shapire available here: 1, 2, 3

 

Code from class

Matlab code for the committee machines.

 

(Homework Assignment #4)


Week 16: April 27-May 1, 2015

 

Topics

Committee machines

Support vector machines

 

Reading material

Textbook: Combining Models (Chapter 14)

Textbook: Sparse kernel machines (Chapter 7)

 


Week 17: May 4-8, 2015

 

Final Exam, Thursday, 2:45pm

 


Last updated: 05/01/2015 04:41 PM