Schedule

The following schedule is possibly optimistic, but is definitely tentative and subject to revision, depending on how we progress. Ideally, we'll touch on all of the topics given here, but we may dwell longer on some, in the interest of improved understanding, at the cost of neglecting others.

Section 0: Introduction and Review

Tue, 8/20

Introduction; administrivia; description of the problem.

Thu, 8/22

Review of probability and statistics I: axiomatic probability; random variables; conditional probability; independence (Happy full moon!)

Tue, 8/27

Review of probability and statistics II: Bayes' rule; estimation; maximum likelihood and maximum a-posteriori

Section I: Supervised Learning - Classification and Regression

Thu, 8/29

Generative and discriminative models; decision theory; risk and cost; Gaussian classification
Homework 1 due

Tue, 9/3

Categorical data: multinomial models; independence; naïve Bayes; intro to graphical models

Thu, 9/5

Continuous data: linear regression
Reading/Discussion

Tue, 9/10

Linear classification: linear machines; logistic regression

Thu, 9/12

Decision trees
Homework 2 due

Tue, 9/17

Learning theory: the PAC model

Thu, 9/19

Instance based methods: nearest-neighbor, kernel methods
Reading/Discussion

Tue, 9/24

Support Vector Machines

Thu, 9/26

Ensemble methods: meta-learning; boosting
Homework 3 due

Section II: Representation, Relational Learning, and Deductive Learning

Tue, 10/1

Rule Induction: FOIL, a-priori

Thu, 10/3

Inductive Logic Programming

Tue, 10/8

Genetic programming
Reading/Discussion

Thu, 10/10

Fall Break - no class. I won't be here. You shouldn't either.

Section III: Unsupervised/Semi-supervised Learning

Tue, 10/15

Clustering I: Similarity-based clustering; K-means

Thu, 10/17

Midterm Exam

Tue, 10/22

Clustering II: Expectation-Maximization

Thu, 10/24

Learning in Bayesian Networks
Final Project Proposal due

Section IV: Time-series Analysis and Reinforcement Learning

Tue, 10/29

Discrete Models/DFAs; Angulin's algorithm

Thu, 10/31

Markov Chains
Reading/Discussion

Tue, 11/5

Hidden Markov Models

Thu, 11/7

Stochastic CFGs
Homework 4 due

Tue, 11/12

Markov Decision Processes

Thu, 11/14

Reinforcement Learning I
Reading/Discussion

Tue, 11/19

RL II (Another full moon...)

Thu, 11/21

POMPDPs

Tue, 11/26

Project presentations/special topics

Thu, 11/28

Happy Thanksgiving! No class.

Tue, 12/3

Project presentations/special topics

Thu, 12/5

Project presentations/special topics
Final Project Report Due

Final Exam Week

Final to be scheduled.