DOI QR코드

DOI QR Code

소량자료를 위한 베이지안 다중 변환점 모형

Bayesian Multiple Change-Point for Small Data

  • 전수영 (고려대학교 정보통계학과) ;
  • Cheon, Soo-Young (Department of Informational Statistics, Korea University) ;
  • Yu, Wenxing (Department of Economics and Statistics, Korea University)
  • 투고 : 2011.12.16
  • 심사 : 2012.01.20
  • 발행 : 2012.03.31

초록

다중 변환점(multiple change-point) 추론에 있어 소량자료에 관한 연구는 많지 않다. 본 논문에서는 소량 자료의 다중 변환점 추정을 위해 베이지안 비중심(noncentral) t 분포 변환점 모형을 제안하고, 제안된 모형 추론을 위해 메트로폴리스-해스팅스를 포함한 깁스 샘플링(Metropolis-Hastings-Within-Gibbs sampling) 알고리즘을 이용하였다. 모의실험 및 태풍 발생 수의 실증 분석결과는 제안된 모형과 알고리즘의 우수성을 보여 준다.

Bayesian methods have been recently used to identify multiple change-points. However, the studies for small data are limited. This paper suggests the Bayesian noncentral t distribution change-point model for small data, and applies the Metropolis-Hastings-within-Gibbs Sampling algorithm to the proposed model. Numerical results of simulation and real data show the performance of the new model in terms of the quality of the resulting estimation of the numbers and positions of change-points for small data.

키워드

참고문헌

  1. 설동일(2010). 지구온난화와태풍의변화경향, <한국항해항만학회학술발표>, 1, 238-239.
  2. Barry, D. and Hartigan, J. A. (1993). A Bayesian analysis for change-point problems, Journal of the American Statistical Association, 88, 309-319. https://doi.org/10.2307/2290726
  3. Carlin, B. P., Gelfand, A. E. and Smith, A. F. M. (1992). Hierarchical Bayesian analysis of change point problem, Applied statistics, 41, 389-405. https://doi.org/10.2307/2347570
  4. Chen, M. H. and Schmeiser, B. W. (1998). Towards black-box sampling, Journal of Computational and Graphical Statistics, 7, 1-22.
  5. Cheon, S. and Kim, J. (2010). Multiple change-point detection of multivariate mean vectors with the Bayesian approach, Computational Statistics & Data Analysis, 54, 406-415. https://doi.org/10.1016/j.csda.2009.09.003
  6. Chib, S. (1998). Estimation and comparison of multiple change-point models, Journal of Econometrics, 86, 221-241. https://doi.org/10.1016/S0304-4076(97)00115-2
  7. Geman, S. and Geman, D. (1984). Stochastic relaxation, Gibbs distributions, and the Bayesian restoration of images, IEEE Transactions on Pattern Analysis and Machine Intelligence, 6, 721-741. https://doi.org/10.1109/TPAMI.1984.4767596
  8. Hastings, W. K. (1970). Monte Carlo Sampling Methods using Markov Chains and their applications, Biometrika, 57, 97-109. https://doi.org/10.1093/biomet/57.1.97
  9. Hinkley, D. V. (1970). Inference about the change-point in a sequence of random variables, Biometrika, 57, 1-17. https://doi.org/10.1093/biomet/57.1.1
  10. Kim, J. and Cheon, S. (2010). Bayesian multiple change-point estimation with annealing stochastic approximation Monte Carlo, Computational Statistics, 25, 215-239. https://doi.org/10.1007/s00180-009-0172-x
  11. Metropolis, N., Rosenbluth, A. W., Rosenbluth, M. N., Teller, A. H. and Teller, E. (1953). Equations of state calculations by fast computing machines, Journal of Chemical Physics, 21, 1087-1091. https://doi.org/10.1063/1.1699114
  12. Muller, P. (1991). A generic approach to posterior integration and Gibbs sampling, Technical Report, Purdue University, West Lafayette IN.
  13. Muller, P. (1992). Alternatives to the Gibbs Sampling Scheme, Duke University.
  14. Smith, A. F. M. (1975). A Bayesian approach to inference about a change-point in a sequence of random variables, Biometrika, 62, 407-416. https://doi.org/10.1093/biomet/62.2.407
  15. Sugi, M., Noda, A. and Sato, N. (2002). Influence of the global warming on tropical cyclone climatology: An experiment with the JMA global model, Journal of the Meteorological Society of Japan, 80, 249-272. https://doi.org/10.2151/jmsj.80.249
  16. Venter, J. H. and Steel, S. J. (1996). Finding multiple abrupt change points, Computational Statistics & Data Analysis, 22, 481-504. https://doi.org/10.1016/0167-9473(96)00007-2

피인용 문헌

  1. Bayesian Multiple Change-Point Estimation of Multivariate Mean Vectors for Small Data vol.25, pp.6, 2012, https://doi.org/10.5351/KJAS.2012.25.6.999