Institute of Applied Physics and Computational Mathematics, Beijing, P. R. China.
Jian-Sheng Wang (National University of Singapore)
This workshop aims to give an introduction to the Monte Carlo method. The Monte Carlo method was initially developed during the 1940s when the electronic computer was just first available. The Monte Carlo method solves problems that are intractable by analytic or other means. The use of Monte Carlo methods spans many areas, such as phase transitions in condensed matter physics, modeling of biological molecules, fluid flow, neutron diffusion in reactor, statistical Bayesian analysis, optimization, image processing, bioinformatics, etc.
In these lectures, we begin with the basic principles of Monte Carlo method, introducing in an elementary way the generation of random numbers, concept of Markov chains, Metropolis algorithm, and analysis of Monte Carlo data. We introduce briefly the quasi-Monte Carlo, and a little bit quantum Monte Carlo in physics. We demonstrate the use of Monte Carlo method from applications in statistical physics, statistics, and other fields. In the later part of the lectures, we'll present more advanced topics such as efficient simulation methods (simulated tempering, multicanonical methods, flat-histogram methd, cluster algorithms). We end the workshop with a few invited review talks on Monte Carlo methods.
Graduate students and junior researchers are most suited for this workshop. Contributed short talks are welcome.
Bo Zheng (Zhejiang University), Nonequilibrium dynamics in statistical physics
Junni Zhang (Beijing University), Markov chain Monte Carlo in statistics - recent advances
Pei Lucheng (Atomic Energy Institute), Monte Carlo method and its characteristics
The language of the lecture presentation will be in Chinese. The workshop will start from Monday 3 November at 8:30am. Mini-presentations contributed by the participants will be organized during the workshop.
Some details of the workshop are available at http://phyweb.physics.nus.edu.sg/~phywjs/BeijingWorkshop.html
Jian-Sheng Wang graduated from Jilin University with a B.Sc in 1982. He joined the CUSPEA programme in 1982 and obtained Ph.D. (1987) in Physics from Carnegie-Mellon University. He had postdoctoral positions at Rutgers University (USA), HLRZ Juelich and University of Mainz (Germany). His first job was at Hong Kong Baptist College from 1991 to 1993. Since 1993, he has been at National University of Singapore. He was in the Department of Computational Science, and now is a professor in the Department of Physics. He is a fellow of the American Physical Society (2005) for his outstanding contributions to the development of novel computer simulation algorithms and for their use in the study of phase transitions and critical phenomena. Dr. Jian-Sheng Wang's research field is in Monte Carlo method in statistical physics (with applications in critical phenomena and phase transitions, spin-glasses, nonequilibrium systems known as driven diffusive systems, random sequential adsorption, and research for efficient algorithms). He is well-known for the development of the cluster algorithm by the name Swendsen-Wang algorithm for simulation of Ising and Potts models. His current interests move to quantum thermal transport, nonequilibrium Green's function methods, and near-field radiative heat transfer.
This pdf file contains a registration form and additional information concerning winter workshops.
On Monday we start at 8:30. For the rest of week, we have morning sessions 9:00 to 12:00, and afternoon sessions 2:30 to 5:30, with some breaks.
J.-S. Wang's Lecture notes (slightly revised Sep 2020, powerpoint):
0 (schedule, books),
1 (introduction),
2 (probability),
3 (random number),
4 (integral by MC),
5 (Quasi Monte Carlo),
6 (Markov chain),
7 (Metropolis algorithm),
8 (applications),
9 (errors and convergence),
10 (quantum MC),
11 (cluster algorithm),
12 (reweighting),
13 (extended ensemble),
14 (TMMC, Wang-Landau),
15 (not covered),
Junni Zhang's Lecture notes: talk_IAPCM.pdf.
Bo Zheng's Lecture notes: beijing_1-3.doc, beijing_4.ppt, beijing_5-6.doc, econophys.ppt.
rand.c: three different random number generators.
ising_Metropolis.c: a simple Ising model simulation.
sw2d.f: Swendsen-Wang algorithm, mainly for timing (poor quality random number generator used).
Wolff.c: Wolff's single cluster for 2D Ising model.
sw-oner-gg.c: compute tau for Metropolis, Swendsen-Wang, Sweeny, etc.
tmmc_n_conv.c: transition matrix Monte Carlo/flat-histogram code.
glassibm.f: legacy fortran code for replica Monte Carlo.
QMC.tar.gz: diffusion quantum Monte Carlo and Hartree-Fock, a tarred and gzipped file.
The PowerPoint slides and the codes are copyright of respective authors (2003).
Please contact Xijun Yu at yuxj@iapcm.ac.cn or Jian-Sheng Wang at phywjs@nus.edu.sg.