Markov Chain Monte Carlo: algorithms for multimodal distribution and their applications

Yukito iba

Presentation information

[J] Oral

M (Multidisciplinary and Interdisciplinary) » M-GI General Geosciences, Information Geosciences & Simulations

[M-GI33] Data-driven geosciences

Mon. May 27, 2019 3:30 PM - 5:00 PM A08 (TOKYO BAY MAKUHARI HALL)

convener:Tatsu Kuwatani(Japan Agency for Marine-Earth Science and Technology), Hiromichi Nagao(Earthquake Research Institute, The University of Tokyo), Kenta Ueki(Japan Agency for Marine-Earth Science and Technology), Masayuki Kano(Graduate school of science, Tohoku University), Chairperson:Tatsu Kuwatani(JAMSTEC), Suguru Yabe(JAMSTEC)

3:30 PM - 4:00 PM

[MGI33-06] Markov Chain Monte Carlo: algorithms for multimodal distribution and their applications

★Invited Papers

*Yukito iba¹ (1.The Institute of Statistical Mathematics)

Keywords:Markov Chain Monte Carlo, Replica Exchange Monte Carlo (Parallel Tempering), Bayesian modeling, rare event sampling

As introduction, history and software of Markov chain Monte Carlo is briefly discussed. Then, replica exchange method (parallel tempering) is explained as a useful tool for sampling multimodal distributions. It is also stressed that applications of Markov chain Monte Carlo are not rescricted to statistical physics or Bayesian computaion; it can deal with a variety of problems related to high dimensional probability, such as counting the number of discrete structures, sampling rare orbits in nonlinear dynamical systems, statistical testing under complicated null hypothesis.