Statistical Modeling II

• This is the GitHub page of course SDS 383D. All the exercises, solutions, and code will be posted.
• Solution to every exercise is edited by blue font in each section.

Section1: Preliminaries

Solutions are available in the folder section01 with file name sds-383d-section01.pdf file.

• 1.1 Exchangeability and de Finetti's theorem

• 1.2 The exponential family of distributions

• 1.3 Multivariate normal distribution

• 1.4 Frequentist estimation and uncertainty quantification

Section2: Bayesian inference in Gaussian models

Solutions are available in the folder section02 with file name sds-383d-section02.pdf file.

• 2.1 Bayesian inference in a simple Gaussian model

• 2.2 Bayesian inference in a multivariate Gaussian model

• 2.3 A Gaussian linear model

• 2.4 A hierarchical Gaussian linear model

Section3: Bayesian Generalized Linear Models

Solutions are available in the folder section03 with file name sds-383d-section03.pdf file.

• 3.1 Modeling non-Gaussian observations

Section4: Gaussian processes

Solutions are available in the folder section04 with file name sds-383d-section04.pdf file.

• 4.1 Non-linear functions

• 4.2 Model selection

• 4.3 Beyond regression: non-conjugate likelihoods