CS 6347: Statistical Methods in AI and ML

Spring 2017

Where: ECSN Building 2.203
When: MW, 11:30am-12:45pm

Instructor: Nicholas Ruozzi
Office Hours: Tuesday/Friday 10am-11am and by appointment in ECSS 3.409

TA: Dhruv Patel
Office Hours: M/W 4pm - 5pm in ECSS 2.104A1

Grading: problem sets (70%), final project (25%), class participation & extra credit (5%)
Attendance is MANDATORY. The instructor reserves the right to lower final grades as a result of poor attendance.

Prerequisites: some familiarity with basic probability, linear algebra, and introductory machine learning (helpful, but not required).

Week	Dates	Topic	Readings
1	Jan. 9 & 11	Introduction & Basic Probability Bayesian Networks	K&F: Ch. 1 & 2 Basic Probability BN Notes
2	Jan. 18	More BNs: D-separation	K&F: Ch. 3 Octave (free version of MATLAB)
3	Jan. 23 & 25	Markov Random Fields	K&F: Ch. 4 & Ch. 9 Darwiche: Ch. 9 MRF Notes
4	Jan. 30 & Feb. 1	Variable Elimination & BP Approx. MAP Estimation MAP LP	K&F: 13.1-13.5, A.5.3 Boyd: Ch. 5.1-5.5
5	Feb. 6 & 8	More Approximate MAP Variational Methods	Approximate MAP Notes K&F 11.1-11.2, 11.5 Sections 1-3 of this paper
6	Feb. 13 & 15	Intro to Sampling Markov Chain Monte Carlo	K&F 12.1-12.3
7	Feb. 20 & 22	Intro to Machine Learning Maximum Likelihood for Bayesian Networks	K&F: 17.1-17.4
8	Feb. 27 & Mar. 1	MLE for Log-Linear Models	K&F: 20.1-20.5
9	Mar. 6 & Mar. 8	Alternatives to MLE Expectation Maximization
10	Mar. 20 & Mar. 22	Hidden Markov Models Structure Learning	K&F: 19.1-19.2 Box 17.E
11	Mar. 27 & Mar. 29	LDA Exponential Families and EP	K&F: 20.6
12	Apr. 3 & Apr. 5	Free Energy Approximations Graph Cuts and Efficient MAP inference	Ising Models General Models

All problem sets will be available on the eLearning site and are to be turned in there. See the homework guidelines below for the homework policies.

This semster, online notes in book form will (hopefully) be available for each lecture. In addition, the following textbook is suggested:

• Probabilistic Graphical Models: Principles and Techniques, by Daphne Koller and Nir Friedman.

Other references that may be helpful:

• Modeling and Reasoning with Bayesian Networks, by Adnan Darwiche.
• Machine Learning: a Probabilistic Perspective, by Kevin Murphy.

We expect you to try solving each problem set on your own. However, when being stuck on a problem, I encourage you to collaborate with other students in the class, subject to the following rules:

1. You may discuss a problem with any student in this class, and work together on solving it. This can involve brainstorming and verbally discussing the problem, going together through possible solutions, but should not involve one student telling another a complete solution.
2. Once you solve the homework, you must write up your solutions on your own, without looking at other people's write-ups or giving your write-up to others.
3. In your solution for each problem, you must write down the names of any person with whom you discussed it. This will not affect your grade.
4. Do not consult solution manuals or other people's solutions from similar courses - ask the course staff, we are here to help!

Late homeworks will NOT be accepted except in extreme circumstances or those permitted by university policy (e.g., a religious holiday). All such extensions MUST be cleared in advance of the due date.