1 / 29

Bayes Factor

Bayes Factor. Based on Han and Carlin (2001, JASA). Model Selection. Data: y : finite set of competing models : a distinct unknown parameter vector of dimension n j corresponding to the j th model Prior : all possible values for

dysis
Télécharger la présentation

Bayes Factor

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Bayes Factor Based on Han and Carlin (2001, JASA)

  2. Model Selection • Data: y • : finite set of competing models • : a distinct unknown parameter vector of dimension nj corresponding to the jth model • Prior • : all possible values for • : collection of all model specific

  3. Model Selection • Posterior probability • A single “best” model • Model averaging • Bayes factor: Choice between two models

  4. Estimating Marginal Likelihood • Marginal likelihood • Estimation • Ordinary Monte Carlo sampling • Difficult to implement for high-dimensional models • MCMC does not provide estimate of marginal likelihood directly • Include model indicator as a parameter in sampling • Product space search by Gibbs sampling • Metropolis-Hastings • Reversible Jump MCMC

  5. Product space search • Carlin and Chib (1995, JRSSB) • Data likelihood of Model j • Prior of model j • Assumption: • M is merely an indicator of which is relevant to y • Y is independent of given the model indicator M • Proper priors are required • Prior independence among given M

  6. Product space search • The sampler operates over the product space • Marginal likelihood • Remark:

  7. Search by Gibbs sampler

  8. Bayes Factor • Provided the sampling chain for the model indicator mixes well, the posterior probability of model j can be estimated by • Bayes factor is estimated by

  9. Choice of prior probability • In general, can be chosen arbitrarily • Its effect is divided out in the estimate of Bayes factor • Often, they are chosen so that the algorithm visits each model in roughly equal proportion • Allows more accurate estimate of Bayes factor • Preliminary runs are needed to select computationally efficient values

  10. More remarks • Performance of this method is optimized when the pseudo-priors match the corresponding model specific priors as nearly as possible • Draw back of the method • Draw must be made from each pseudo prior at each iteration to produce acceptably accurate results • If a large number of models are considered, the method becomes impractical

  11. Metropolized product space search • Dellaportas P., Forster J.J., Ntzoufras I. (2002). On Bayesian Model and Variable Selection Using MCMC.  Statistics and Computing, 12, 27-36. • A hybrid Gibbs-Metropolis strategy • Model selection step is based on a proposal moving between models

  12. Metropolized Carlin and Chib (MCC)

  13. Advantage • Only needs to sample from the pseudo prior for the proposed model

  14. Reversible jump MCMC • Green (1995, Biometrika) • This method operates on the union space • It generates a Markov chain that can jump between models with parameter spaces of different dimensions

  15. RJMCMC

  16. Using Partial Analytic Structure (PAS) • Godsill (2001, JCGS) • Similar setup as in the CC method, but allows parameters to be shared between different models. • Avoids dimension matching

  17. PAS

  18. Marginal likelihood estimation (Chib 1995)

  19. Marginal likelihood estimation (Chib 1995) • Let When all full conditional distributions for the parameters are in closed form

  20. Chib (1995) • The first three terms on the right side are available in close form • The last term on the right side can be estimated from Gibbs steps

  21. Chib and Jeliazkov (2001) • Estimation of the last term requires knowing the normalizing constant • Not applicable to Metropolis-Hastings • Let the acceptance probability be

  22. Chib and Jeliazkov (2001)

  23. Example: linear regression model

  24. Two models

  25. Priors

  26. For MCC • h(1,1)=h(1,2)=h(2,1)=h(2,2)=0.5

  27. For RJMCMC • Dimension matching is automatically satisfied • Due to similarities between two models

  28. Chib’s method • Two block gibbs sampler • (regression coefficients, variance)

  29. Results • By numerical integration, the true Bayes factor should be 4862 in favor of Model 2

More Related