---

Prog. Theor. Phys. Supplement No.138 (2000) pp. 462-463

Identification of Metastable States in Finite Temperature Simulations: A Self-Organization Approach

Yukito Iba^1,* and Koji Hukushima^2,**

¹The Institute of Statistical Mathematics, Tokyo 106-8569, Japan
² Institute for Solid State Physics, University of Tokyo, Tokyo 106-8666, Japan

(Received October 11, 1999)

Abstract:

How can we identify “metastable states” from a simulation of a finite size system at finite temperatures? Here we propose a self-organization approach to this problem based on a finite mixture model and Expectation-Maximization (EM) algorithm and apply it to the analysis of Monte Carlo simulation of Ising spin glass models.

URL : http://ptp.ipap.jp/link?PTPS/138/462/
DOI : 10.1143/PTPS.138.462

^*E-mail address: iba@ism.ac.jp
^**E-mail address: hukusima@chiral.issp.u-tokyo.ac.jp

[ Full Text PDF : FREE ACCESS (74K) ] Citation: Plain Text BiBTeX EndNote(RIS) RSS

---

References:

1. D. M. Titterington, A. F. M. Smith and U. E. Makov, Statistical Analysis of Finite Mixture Distributions (John Wiley and Sons, Chichester, 1985).

---

[PTP HOME] [BROWSE] [SEARCH] [SUBSCRIPTION] [SUBMISSION] [REGISTRATION]

Copyright © Progress of Theoretical Physics 2013 All rights reserved.