2019年度 総合数理特論B (ベイズ統計学の入門・大学院理学研究科)



Instructor


Texts


Schedule

Week 16 (2/6): Semiparametric Bayesian copula estimation Week 15 (2/4): Linear mixed-effect models Week 14 (1/30): Data augmentation Week 13 (1/28): Metropolis-Hastings algorithm Week 12 (1/23): Metropolis algorithm Week 11 (1/21): Variable selection and sparse linear models Week 10 (1/16): Hierarchical modeling (cont'd) and start Bayesian linear regression Week 9 (1/9): (Missing data and imputation,) Hierarchical modeling Week 8 (1/7): Canceled Week 7 (12/24): Multivariate normal model Week 6 (12/19): Introduction to MCMC (Gibbs sampling) Week 5 (12/17): Normal distribution Week 4 (12/12): Objective prior, Monte Carlo approximation Week 3 (12/10): Credible interval, Count data Week 2 (12/5): Binary data, Conjugate prior, Posterior predictive density Week 1 (12/3): Introduction

Evaluation


Outline

  1. Probability and Bayes' formula (12/3)
  2. Inference for one-parameter models (12/5, 12/10)
  3. Monte Carlo approximation (12/12)
  4. Inference for the normal distribution (12/17)
  5. Markov chain Monte Carlo method (Gibbs sampling: 12/19; MH algorithm: 1/23, 28)
  6. Multivariate normal model (12/24), Missing data (1/9)
  7. Hierarchical models (1/9,1/16)
  8. Linear regression models and Bayesian lasso (1/16, 1/21)
  9. Data augmentation and for generalized linear models (1/30)
  10. Linear mixed-effect models (2/4)
  11. Semiparametric copula estimation (2/6)