Overview
Course name: Monte Carlo Methods and Bayesian Computation
Instructor:
Date: Spring, 2021
Course Description:
-
Simulation and the Monte Carlo Method, 3rd Edition, Rubinstein and Kroese, Wiley
-
Bayesian Methods for Data Analysis, 3rd Edition, Carlin and Louis, Chapman & Hall
-
Monte Carlo Statistical Methods, 2nd Edition, Robert and Casella, Springer
Objectives:
-
Bayesian framework of inference
-
Markov Chain Monte Carlo methods (MCMC) to estimate Bayesian posterior distributions (Gibbs sampler, Metropolis-Hastings algorithm, Hamiltonian algorithm)
-
Sensitivity analyses and the splitting method for analyzing difficult estimation problems will also be discussed.
-
Monte Carlo methods, including Monte Carlo integration (classic, importance sampling, Laplacian, saddlepoint approximation)
-
variance reduction techniques (Rao-Blackwellization, Control and antithetic variates)
-
Monte Carlo optimization (EM algorithm).