Talks

Location: Home>Activities>Talks

Irreversible Markov Chains for Particle Systems and Spin Models: Mixing and Dynamical Scaling

04/29 2019
  • Title Irreversible Markov Chains for Particle Systems and Spin Models: Mixing and Dynamical Scaling
  • Speaker
  • Date
  • Venue
  • Abstract

    CAS Key Laboratory of Theoretical Physics

    Institute of Theoretical Physics

    Chinese Academy of Sciences

    Seminar

    Title

    题目

    Irreversible Markov Chains for Particle Systems and Spin Models: Mixing and Dynamical Scaling

    Speaker

    报告人

    Ze Lei (雷泽)                                        

    Affiliation

    所在单位

    Laboratoire de Physique Statistique, école Normale Supérieure de Paris

    Date

    日期

    2019/04/29 上午10点
     

    Venue

    地点

    6520 ITP South Building

    Abstract

    摘要

    This seminar presents irreversible Markov chains in particle and spin systems, focusing on the difference between the scales of "mixing time" and "relaxation time" in the dynamics. In particular, the mechanism of an "event-chain" algorithm in one-dimensional hard-sphere model is related to a coupon-collector problem; in two-dimensional XY model, a slow vortex-pair mode and a fast spin-wave mode are spotted. As a generalization, a new optimization ("factor field") is proposed based on the previous studies, and a super-diffusive behavior is identified in the present algorithm.

    Contact Person

    所内联系人

    金瑜亮