Schedule

A schedule of topics and readings is provided below. Each week will cover a single topic or group of topics. You should treat the readings as a reference and as a more detailed exposition of the topics discussed in lecture. Consult the readings when you want to know more or want a slightly different approach to explaining a particular topic.

Monday lectures will typically introduce the core theory for the week’s topic while Wednesday lectures will go into greater detail and work through applications. You should make sure to review the readings prior to that week’s lectures with an aim towards completing them before Wednesday’s lecture.

All hyperlinks to papers are either to the published version of the paper (if published) or to the working paper. You should use your university library access to obtain the full version if the article is not open access. Likewise, textbook readings will be to the free/open version of the book (if available) or to the UW-Madison university library eBook.

There is no class during the first week of instruction due to APSA and Labor Day (September 2 and 7). The course begins on Wednesday, September 9.

Week 1: Course Introduction and Review

Wednesday, September 9

Topics:

  • What is a statistical model and what is it good for?
  • Review of regression

Readings:

  • Review: Chapters 1-7; Gelman, Andrew, Jennifer Hill, and Aki Vehtari. 2020. Regression and Other Stories. Cambridge: Cambridge University Press.

Problem Set 1 assigned Wednesday, September 9 – due Wednesday, September 23.


Week 2: Introduction to Likelihood Inference

Monday, September 14 & Wednesday, September 16

Topics:

  • What is a “parametric” model?
  • The likelihood function
  • Maximum likelihood estimation

Readings:


Week 3: Generalized Linear Models

Monday, September 21 & Wednesday, September 23

Topics:

  • Properties of maximum likelihood estimators
  • The generalized linear model framework
  • Binary outcome models

Readings:

Problem Set 1 due Wednesday, September 23. Problem Set 2 assigned Wednesday, September 23 – due Wednesday, October 7.


Week 4: More Likelihood Models

Monday, September 28 & Wednesday, September 30

Topics:

  • Duration models (parametric, semi-parametric (Cox) and non-parametric approaches)
  • Event count models (Poisson regression and quasi-MLE methods)

Readings:


Week 5: Bayesian Inference

Monday, October 5

Topics:

  • Principles of posterior inference
  • How to write a Bayesian model
  • Quantities of interest: posterior mode, posterior mean, credible intervals

Readings:

  • Chapter 9; Gelman, Andrew, Jennifer Hill, and Aki Vehtari. 2020. Regression and Other Stories. Cambridge: Cambridge University Press.
  • Chapters 10-11; Gelman, Andrew, John Carlin, Hal Stern, David Dunson, Aki Vehtari, and Donald Rubin. 2013. Bayesian Data Analysis. 3rd ed. Boca Raton: Chapman and Hall/CRC.

Problem Set 2 due Wednesday, October 7. Problem Set 3 assigned Wednesday, October 7 – due Wednesday, October 21.


Midterm 1 In-Class, Wednesday, October 7th


Week 6: Estimating Bayesian Models

Monday, October 12 & Wednesday, October 14

Topics:

  • Analytically tractable solutions through conjugate priors
  • Markov Chain Monte Carlo methods
  • Introduction to MCMC in Stan

Readings:


Week 7: Multilevel Regression Models

Monday, October 19 & Wednesday, October 21

Topics:

  • Partial pooling through priors
  • Methods for model validation
  • Multilevel regression and post-stratification (MRP)

Readings:

Problem Set 3 due Wednesday, October 21. Problem Set 4 assigned Wednesday, October 21 – due Wednesday, November 4.

Replication project memo due Wednesday, October 21


Week 8: Mixture Models and the EM Algorithm

Monday, October 26 & Wednesday, October 28

Topics:

  • Exploratory data analysis and clustering models
  • MLE and MAP estimation via the “Expectation-Maximization” algorithm

Readings:

  • Chapters 13 and 22; Gelman, Andrew, John Carlin, Hal Stern, David Dunson, Aki Vehtari, and Donald Rubin. 2013. Bayesian Data Analysis. 3rd ed. Boca Raton: Chapman and Hall/CRC.

Week 9: Item Response Theory and Ideal Point Models

Monday, November 2

Topics:

  • Latent trait models
  • “Ideal point” models in legislative voting

Readings:

Problem Set 4 due Wednesday, November 4.

Midterm 2 In-Class, Wednesday, November 4th


Week 10: Regularization and Model Selection

Monday, November 9 & Wednesday, November 11

Topics:

  • Variable selection and penalized regression (Ridge, LASSO)
  • Cross-fitting and out-of-sample validation

Readings:


Week 11: Flexible Regression – Trees and Forests

Monday, November 16 & Wednesday, November 18

Topics:

  • Regression trees
  • Aggregation via sums of trees + regularization via BART
  • Bagging, boosting, random forests

Readings:


Week 12: Flexible Regression – Kernels and Gaussian Processes

Monday, November 23

Thanksgiving recess runs November 26-29. Class meets as usual on Monday this week. Wednesday is cancelled.

Topics:

  • Infinite dimensional function spaces
  • Kernel ridge regression
  • Gaussian process regression

Readings:


Week 13: Causal Inference with Flexible Regression

Monday, November 30 & Wednesday, December 2

Topics:

  • Efficient semiparametric estimation of statistical functionals
  • Influence functions
  • “Doubly-robust”/“Neyman-orthogonal” estimators

Readings:


Poster Session Friday, December 4th


Week 14: Working with “Big” Datasets

Monday, December 7 & Wednesday, December 9

Topics:

  • Estimation with out-of-memory data
  • Stochastic gradient descent
  • Lossless compression in regression

Readings: