Journal of the Korean Statistical Society

The Korean Statistical Society (한국통계학회)
권호 표지

Nonparametric Bayesian Multiple Change Point Problems

  • Kim, Chansoo (Research Institute of Computer, Information and Communication, Pusan National University) ;
  • Younshik Chung (Department of Statistics, Pusan National University)
  • Published : 2002.03.01
Download PDF
⟨ Previous Next ⟩

Abstract

Since changepoint identification is important in many data analysis problem, we wish to make inference about the locations of one or more changepoints of the sequence. We consider the Bayesian nonparameteric inference for multiple changepoint problem using a Bayesian segmentation procedure proposed by Yang and Kuo (2000). A mixture of products of Dirichlet process is used as a prior distribution. To decide whether there exists a single change or not, our approach depends on nonparametric Bayesian Schwartz information criterion at each step. We discuss how to choose the precision parameter (total mass parameter) in nonparametric setting and show that the discreteness of the Dirichlet process prior can ha17e a large effect on the nonparametric Bayesian Schwartz information criterion and leads to conclusions that are very different results from reasonable parametric model. One example is proposed to show this effect.

Keywords

References

  1. The Annals of Statistics v.2 Mixtures of Dirichlet processes with applications to nonparametric problems Antoniak, C. E. https://doi.org/10.1214/aos/1176342871
  2. Applied Statistics v.41 Hierarchical Bayesian analysis of changepoint problems Carlin, B. P.; Gelfand, A. E.; Smith, A. F. M. https://doi.org/10.2307/2347570
  3. The Annals of Mathematical Statistics v.35 Estimating the correct mean of a normal distribution which is subjected to a change in time Chernoff, H.; Zacks, S. https://doi.org/10.1214/aoms/1177700517
  4. Estimating the Means of Several Normal Populations by Nonparametric Estimation of the Distribution of the Means, unoublished dissertation Escobar, M. D.
  5. Journal of the American Statistical Association v.89 Estimating normal means with a Dirichlet process prior Escobar, M. D. https://doi.org/10.2307/2291223
  6. Journal of the American Statistical Association v.90 Bayesian density estimation and inference using mixtures Escobar, M. D.; West, M. https://doi.org/10.2307/2291069
  7. The Annals of Statistics v.1 Bayesian analysis of some nonparametric problems Ferguson, T. S. https://doi.org/10.1214/aos/1176342360
  8. Journal of the American Statistical Association v.85 Sampling-Based approaches to calculating marginal densities Gelfans, A. E.; Smith, A. F. M. https://doi.org/10.2307/2289776
  9. Biometrika v.67 Inference about the changepoint from cumulative sum tests Hinkley, D. V. https://doi.org/10.1093/biomet/67.1.79
  10. SIAM Journal of Sciuntific and Statistical Computing v.7 Computations of mixtures of Dirichlet processes Kuo, L. https://doi.org/10.1137/0907004
  11. Communications in Statistics : Simulation and Computation v.23 Estimating normal means with a conjugate style Dirichlet process prior MacEachern, S. N. https://doi.org/10.1080/03610919408813196
  12. Bayesian Statistics v.5 Bayesian hierarchical nonparametric inference for change-point problems Mira, A.; Petrone, S.
  13. Biometrika v.67 A simple cummulative sum type statistic for the change-point problem with zero-one observations Petitt, A. N. https://doi.org/10.1093/biomet/67.1.79
  14. Biometrika v.62 A Bayesian approach to inference about a changepoint in a sequence of random variables Smith, A. F. M. https://doi.org/10.1093/biomet/62.2.407
  15. TR0002, Department of Statistics Bayesian binary segmentation procedure for homogeneous Poisson process with multiple change positions Yang, T.; Kuo, L.
  16. In Recent Advances in Statistics : Herman Chernoff Festschrift Survey of classical and Bayesian approaches to the change point problem : fixed sample and sequential procedures of testing and estimation Zacks. S.