MonteCarloMethodsStats

Repository for Monte Carlo Methods in Statistics -- includes code for sampling (rejection, metropolis-hastings, Gibbs) and for Bayesian inference + model selection

This repository serves as a personal reference for the SDS386D: Monte Carlo Methods In Statistics course at UT Austin

Overview

• Assignment 1: rejection sampling, importance sampling
• Assignment 2: MCMC
• Assignment 3: MCMC, Metropolis-Hastings, Bayesian inference
• Assignment 4: Gibbs sampling, expected value from the predictive density, use of latent variables, Bayesian inference
• Assignment 5: Bayesian inference for a mixture model, Dirichlet prior, use of latent variables, Gibbs sampling with a Metropolis step
• Assignment 6: sequential Monte Carlo: Kalman filter, particle filter
• midterm: data augmentation via slice sampling, Bayesian inference, Metropolis-Hastings, Gibbs sampling with a Metropolis step
• final: Bayesian model selection