STAT 339: Tentative Schedule of Topics

.

The following is a tentative schedule of what we will cover in class when. Check this space regularly for updates.

Last update: 09-18-17

Key to reading materials:

Date Topics Readings Assignments
M 1/30 What is Machine Learning? MLPP 1.1-1.3, BMRL 13.1
  [PART I: Machine Learning Basics (~ 2 weeks)]  
W 2/01 Nearest Neighbor Classification BRML 14.1-14.2
  Evaluating a Classifier    
F 2/03 Linear Regression FCML 1.1-1.4, ISL Ch. 3
M 2/06 (cont’d)   HW0: Coding Warmup due
W 2/08 (cont’d)  
F 2/10 Utility, Loss, and Validation FCML 1.5, ISL Ch. 5
M 2/13 (cont’d)  
  [PART II: Probabilistic Modeling (~ 4 weeks)]  
W 2/15 Probability Fundamentals FCML 2.1-2.6, BRML 1.1, 8.1-8.4
F 2/17 (cont’d)   HW1: Classification and Regression due
M 2/20 (cont’d)    
W 2/22 Likelihood-Based Inference FCML 2.7-2.11, BRML 8.5-8.8
F 2/24 Bias/Variance Decomposition  
M 2/27 Bias and Variance in Regression   HW2a: Probability Basics due
W 3/01 Bayesian Inference Fundamentals FCML 3.1-3.7, BRML 1.2-1.4  
F 3/03 (cont’d)  
M 3/06 (cont’d)  
W 3/08 Naive Bayes Classification FCML 5.1-5.2, BRML 10.1-10.3 HW2b: Likelihood and Bayesian Inference due
F 3/10 (cont’d)   Take-home Midterm Distributed
M 3/13 Bayesian Regression & Model Selection FCML 3.8-3.9, BRML 12.1-12.4 Take-home Midterm Due
W 3/15 Approximate Inference FCML Ch. 4, 9.1-9.4, BRML 27.1-27.4  
F 3/17 (cont’d)  
  SPRING BREAK
M 3/27 Sampling HW3: Bayesian Modeling due
  [PART III: Unsupervised Learning and Latent Variable Models (~3 weeks)]  
W 3/29 Mixture Models and the EM Algorithm FCML Ch.6, BRML 20.1-20.3
F 3/31 (cont’d)  
M 4/03 (cont’d)  
W 4/05 (cont’d)  
F 4/07 Markov Chain Monte Carlo  
M 4/10 (cont’d)  
W 4/12 Belief Networks FCML 3.6, BRML 3.1-3.3
F 4/14 (cont’d)  
M 4/17 Hidden Markov Models BRML 23.1-23.3, 23.5 HW4: Approximate Inference/Clustering due
W 4/19 (cont’d)    
F 4/21 (cont’d)  
  [PART IV: Nonparametric Models (2 weeks)]  
M 4/24 Gaussian Processes FCML Ch.8, BRML 19.1-19.2, 19.5
W 4/26 (cont’d)  
F 4/28 (cont’d)  
M 5/1 Infinite Mixture Models FCML Ch. 10 HW5: Graphical Models/Hidden Markov Models due
W 5/3 (cont’d)  
F 5/5 Infinite Hidden Markov Models notes
F 5/12 9-11 am GROUP PRESENTATIONS Project Writeups due