MCMC and SMC in Dynamic Models and Stochastic Volatility Models
III Summer School – Technical University of Catalonia
Barcelona, Spain
Course material (PDF FILE)
Textbook
Most of the material from lectures one through six are based on my book entitled MCMC: Stochastic Simulation for Bayesian Inference, co-authored by D. Gamerman and published by Chapman&Hall/CRC in 2006. R code for many of the book’s worked examples can be found at www.dme.ufrj.br/mcmc.
Course outline
Lecture 1: Bayesian inference
Lecture 2: Bayesian model criticism
Lecture 3: Monte Carlo (MC) methods
Lecture 4: Markov chain Monte Carlo (MCMC) methods
Lecture 5: Dynamic linear models (DLM)
Lecture 6: Nonnormal, nonlinear dynamic models
Lecture 7: Stochastic volatility models as dynamic models
Lecture 8: Sequential Monte Carlo (SMC) methods
Lecture 9: SMC with parameter learning
Lecture 10: SMC in stochastic volatility models
R code
- Lecture 1: (boxtiao.R)
- Lecture 3: (montecarlo.R) (simplemontecarlo.R) (rejection.R) (SIR.R) (MC-classroom.R) (REJECTION-classroom.R) (SIR-classroom.R)
- Lecture 4: (markovchain.R) (poisson-changepoint.R) (metropolis-hastings-mixtures.R) (AR1.R) (AR1-drift.R) (GARCH.R)
- Lecture 5: (dlm.R) (dlm-ffbs.R)
- Lecture 6: (bernoulli-regression.R) (nonlineardynamicmodel.R)
- Lecture 7: (sv-ar1.R) (sv-ar1-routines.R)
- Lecture 8: (dlm-smc.R) (bootstrapfilter-stepbystep.R) (LW-stepbystep.R) (nonlineardynamicmodel-smc.R) (dlm-smc-smoothing.R)
- Lecture 9: (dlm-smc-learningsig2-LW.R) (dlm-smc-learningsig2-PL.R) (nonlinearmodel-LW.R)
- Lecture 10: (sv-LW.R)
Other useful links