random cluster dynamics for the ising model is rapidly
play

Random Cluster Dynamics for the Ising model is Rapidly Mixing Heng - PowerPoint PPT Presentation

Feb 25, 2023 •340 likes •1.35k views

Random Cluster Dynamics for the Ising model is Rapidly Mixing Heng Guo Queen Mary, University of London Joint work with Mark Jerrum Oxford Nov 03 2016 Heng Guo (QMUL) Random Cluster 2016/11/03 1 / 41 The model and its dynamics Heng Guo

  1. Random Cluster Dynamics for the Ising model is Rapidly Mixing Heng Guo Queen Mary, University of London Joint work with Mark Jerrum Oxford Nov 03 2016 Heng Guo (QMUL) Random Cluster 2016/11/03 1 / 41

  2. The model and its dynamics Heng Guo (QMUL) Random Cluster 2016/11/03 2 / 41

  3. The random cluster model [Fortuin, Kasteleyn 1969] Parameters 0 ⩽ p ⩽ 1 (edge weight), q ⩾ 0 (cluster weight). Given graph G = ( V , E ) , the measure on subgraph r ⊆ E is defined as π RC ( r ) ∝ p | r | ( 1 − p ) | E \ r | q κ ( r ) , where κ ( r ) is the number of connected components in ( V , r ) . ( 1 − p ) 4 q 4 p 2 ( 1 − p ) 2 q 2 p 4 q Heng Guo (QMUL) Random Cluster 2016/11/03 3 / 41

  4. The random cluster model [Fortuin, Kasteleyn 1969] The partition function (normalizing factor): ∑ p | r | ( 1 − p ) | E \ r | q κ ( r ) . Z RC ( p , q ) = r ⊆ E Equivalent to the Tutte polynomial Z Tutte ( x , y ) : p = 1 − 1 q = ( x − 1 )( y − 1 ) y Heng Guo (QMUL) Random Cluster 2016/11/03 4 / 41

  5. The random cluster model [Fortuin, Kasteleyn 1969] π RC ( r ) ∝ p | r | ( 1 − p ) | E \ r | q κ ( r ) The motivation is to unify: Ising model Potts model Bond percolation Electrical network Heng Guo (QMUL) Random Cluster 2016/11/03 5 / 41

  6. The random cluster model [Fortuin, Kasteleyn 1969] π RC ( r ) ∝ p | r | ( 1 − p ) | E \ r | q κ ( r ) The motivation is to unify: Ising model q = 2 Potts model Bond percolation Electrical network Heng Guo (QMUL) Random Cluster 2016/11/03 5 / 41

  7. The random cluster model [Fortuin, Kasteleyn 1969] π RC ( r ) ∝ p | r | ( 1 − p ) | E \ r | q κ ( r ) The motivation is to unify: Ising model q = 2 Potts model q > 2, integer Bond percolation Electrical network Heng Guo (QMUL) Random Cluster 2016/11/03 5 / 41

  8. The random cluster model [Fortuin, Kasteleyn 1969] π RC ( r ) ∝ p | r | ( 1 − p ) | E \ r | q κ ( r ) The motivation is to unify: Ising model q = 2 Potts model q > 2, integer Bond percolation q = 1 (On K n , Erd˝ os-Rényi random graph) Electrical network Heng Guo (QMUL) Random Cluster 2016/11/03 5 / 41

  9. The random cluster model [Fortuin, Kasteleyn 1969] π RC ( r ) ∝ p | r | ( 1 − p ) | E \ r | q κ ( r ) The motivation is to unify: Ising model q = 2 Potts model q > 2, integer Bond percolation q = 1 (On K n , Erd˝ os-Rényi random graph) q → 0 (Spanning trees if p → 0 and q Electrical network p → 0) Heng Guo (QMUL) Random Cluster 2016/11/03 5 / 41

  10. The random cluster model [Fortuin, Kasteleyn 1969] π RC ( r ) ∝ p | r | ( 1 − p ) | E \ r | q κ ( r ) The motivation is to unify: Ising model q = 2 Potts model q > 2, integer Bond percolation q = 1 (On K n , Erd˝ os-Rényi random graph) q → 0 (Spanning trees if p → 0 and q Electrical network p → 0) Heng Guo (QMUL) Random Cluster 2016/11/03 5 / 41

  11. Glauber dynamics Glauber dynamics (single edge update) P RC (Metropolis): Current state x ⊆ E With prob. 1 / 2 do nothing. (Lazy) 1 Otherwise, choose an edge e u.a.r. 2 { } 1, π RC ( y ) Move to y = x ⊕ { e } with prob. min . 3 π RC ( x ) Detailed balance: π ( x ) P ( x , y ) = π ( y ) P ( y , x ) = min { π ( x ) , π ( y ) } Heng Guo (QMUL) Random Cluster 2016/11/03 6 / 41

  12. Glauber dynamics Glauber dynamics (single edge update) P RC (Metropolis):  { } 1 1, π RC ( y ) 2 m min if | x ⊕ y | = 1;  π RC ( x )     { } P RC ( x , y ) = 1 1, π RC ( x ⊕ { e } ) 1 − e ∈ E min if x = y ; ∑ 2 m π RC ( x )    0 otherwise.   We are interested in the mixing time τ ϵ ( P RC ) : τ ϵ ( P RC ) = min { t : || P t RC ( x 0 , · ) − π || TV ⩽ ϵ } . Heng Guo (QMUL) Random Cluster 2016/11/03 7 / 41

  13. A simple example Let p < 1 / 2. { } 1, π RC ( x ∪ { e } ) min π RC ( x )  p if e is not a cut edge   1 − p = p if e is a cut edge   q ( 1 − p ) Heng Guo (QMUL) Random Cluster 2016/11/03 8 / 41

  14. A simple example Let p < 1 / 2. { } 1, π RC ( x ∪ { e } ) min π RC ( x )  p if e is not a cut edge   1 − p = p if e is a cut edge   q ( 1 − p ) Heng Guo (QMUL) Random Cluster 2016/11/03 8 / 41

  15. A simple example Let p < 1 / 2. { } 1, π RC ( x ∪ { e } ) min π RC ( x )  p if e is not a cut edge   1 − p = p if e is a cut edge   q ( 1 − p ) Heng Guo (QMUL) Random Cluster 2016/11/03 8 / 41

  16. A simple example Let p < 1 / 2. { } 1, π RC ( x ∪ { e } ) min π RC ( x )  p if e is not a cut edge   1 − p = p if e is a cut edge   q ( 1 − p ) Heng Guo (QMUL) Random Cluster 2016/11/03 8 / 41

  17. A simple example Let p < 1 / 2. { } 1, π RC ( x ∪ { e } ) min π RC ( x )  p if e is not a cut edge   1 − p = p if e is a cut edge   q ( 1 − p ) Heng Guo (QMUL) Random Cluster 2016/11/03 8 / 41

  18. Brief History Studied extensively for special graphs, such as the complete graph (mean-field) and the lattice Z 2 . Mean-field: [Gore, Jerrum 1999] [Blanca, Sinclair 2015] Z 2 : [Borgs et al. 1999] [Blanca, Sinclair 2016] [Gheissari, Lubetzky 2016] q > 2: Slow mixing for the complete graph. 0 ⩽ q ⩽ 2: No known fast mixing bound for general graphs. Heng Guo (QMUL) Random Cluster 2016/11/03 9 / 41

  19. Main theorem Theorem For the random cluster model with parameters 0 < p < 1 and q = 2 , τ ϵ ( P RC ) ⩽ 10 n 4 m 2 ( ln π RC ( x 0 ) − 1 + ln ϵ − 1 ) . For q > 2, there exists p such that P RC is slow mixing on complete graphs. [Gore, Jerrum 1999] [Blanca, Sinclair 2015] For q > 2 and 0 < p < 1, it is #BIS-hard to approximate Z RC ( p , q ) . [Goldberg, Jerrum 2012] For 0 ⩽ q < 2, there is no known obstacle. Heng Guo (QMUL) Random Cluster 2016/11/03 10 / 41

  20. Main theorem Theorem For the random cluster model with parameters 0 < p < 1 and q = 2 , τ ϵ ( P RC ) ⩽ 10 n 4 m 2 ( ln π RC ( x 0 ) − 1 + ln ϵ − 1 ) . For q > 2, there exists p such that P RC is slow mixing on complete graphs. [Gore, Jerrum 1999] [Blanca, Sinclair 2015] For q > 2 and 0 < p < 1, it is #BIS-hard to approximate Z RC ( p , q ) . [Goldberg, Jerrum 2012] For 0 ⩽ q < 2, there is no known obstacle. Heng Guo (QMUL) Random Cluster 2016/11/03 10 / 41

  21. Main theorem Theorem For the random cluster model with parameters 0 < p < 1 and q = 2 , τ ϵ ( P RC ) ⩽ 10 n 4 m 2 ( ln π RC ( x 0 ) − 1 + ln ϵ − 1 ) . For q > 2, there exists p such that P RC is slow mixing on complete graphs. [Gore, Jerrum 1999] [Blanca, Sinclair 2015] For q > 2 and 0 < p < 1, it is #BIS-hard to approximate Z RC ( p , q ) . [Goldberg, Jerrum 2012] For 0 ⩽ q < 2, there is no known obstacle. Heng Guo (QMUL) Random Cluster 2016/11/03 10 / 41

  22. Main theorem Theorem For the random cluster model with parameters 0 < p < 1 and q = 2 , τ ϵ ( P RC ) ⩽ 10 n 4 m 2 ( ln π RC ( x 0 ) − 1 + ln ϵ − 1 ) . For q > 2, there exists p such that P RC is slow mixing on complete graphs. [Gore, Jerrum 1999] [Blanca, Sinclair 2015] For q > 2 and 0 < p < 1, it is #BIS-hard to approximate Z RC ( p , q ) . [Goldberg, Jerrum 2012] For 0 ⩽ q < 2, there is no known obstacle. Heng Guo (QMUL) Random Cluster 2016/11/03 10 / 41

  23. Swandsen-Wang algorithm Heng Guo (QMUL) Random Cluster 2016/11/03 11 / 41

  24. Ferromagnetic Ising model [Ising, Lenz 1925] Parameter β > 1. A configuration σ : V → { + , − } . π Ising ( σ ) ∝ β mono ( σ ) = β m − cut ( σ ) Partition function Z Ising ( β ) = ∑ σ β mono ( σ ) β 0 β 2 β 4 Heng Guo (QMUL) Random Cluster 2016/11/03 12 / 41

  25. Equivalence at q = 2 1 Let β = 1 − p . Z Ising ( β ) = β | E | Z RC ( p , 2 ) Heng Guo (QMUL) Random Cluster 2016/11/03 13 / 41

  26. Swendsen-Wang algorithm [Swendsen, Wang 1987] A global Markov chain to sample Ising configurations. Current configuration σ Mark all monochromatic edges under σ as M 1 Remove each edge in M with probability β − 1 (Recall β − 1 = 1 − p ) 2 Assign a random spin to each component of ( V , M ) 3 Practically very fast for the Ising model, but difficult to analyze. Conjectured to be rapidly mixing for all graphs. (Open problem since 90s.) Heng Guo (QMUL) Random Cluster 2016/11/03 14 / 41

  27. Another simple example Activate mono edges 1 Re-randomize mono edges 2 Color components 3 Heng Guo (QMUL) Random Cluster 2016/11/03 15 / 41

  28. Another simple example Activate mono edges 1 Re-randomize mono edges 2 Color components 3 Heng Guo (QMUL) Random Cluster 2016/11/03 15 / 41

  29. Another simple example Activate mono edges 1 Re-randomize mono edges 2 Color components 3 Heng Guo (QMUL) Random Cluster 2016/11/03 15 / 41

  30. Another simple example Activate mono edges 1 Re-randomize mono edges 2 Color components 3 Heng Guo (QMUL) Random Cluster 2016/11/03 15 / 41

  31. Another simple example Activate mono edges 1 Re-randomize mono edges 2 Color components 3 Heng Guo (QMUL) Random Cluster 2016/11/03 15 / 41

  32. Another simple example Activate mono edges 1 Re-randomize mono edges 2 Color components 3 Heng Guo (QMUL) Random Cluster 2016/11/03 15 / 41

Download Presentation
Download Policy: The content available on the website is offered to you 'AS IS' for your personal information and use only. It cannot be commercialized, licensed, or distributed on other websites without prior consent from the author. To download a presentation, simply click this link. If you encounter any difficulties during the download process, it's possible that the publisher has removed the file from their server.

Recommend

Ising model with random boundary condition A. C. D. van Enter # , K. Neto y &amp; , and H. G.

Ising model with random boundary condition A. C. D. van Enter # , K. Neto y &amp; , and H. G.

Equilibrium and Dynamics of Spin Glasses Monte Verita, April 18-23, 2004 Ising model with random boundary condition A. C. D. van Enter # , K. Neto y &amp; , and H. G. Schaap # cn # Institute for Theoretical Physics R.U.G. Groningen, The

473 views • 9 slides
A power-law upper bound on the correlations in the two-dimensional random-field Ising model Ron

A power-law upper bound on the correlations in the two-dimensional random-field Ising model Ron

A power-law upper bound on the correlations in the two-dimensional random-field Ising model Ron Peled, Tel Aviv University Based on a joint work with Michael Aizenman, Princeton ICMP 2018, Montreal Random-field Ising model Standard Ising

447 views • 12 slides
Cluster algorithm for the Ising model Define bond index corresponding to pair of interacting spins

Cluster algorithm for the Ising model Define bond index corresponding to pair of interacting spins

Cluster algorithm for the Ising model Define bond index corresponding to pair of interacting spins Number of bonds N b = dN for a d-dimensional cubic lattice Write the energy of the Ising ferromagnet as Write the partition function as Define

300 views • 14 slides
Quantum Integer Programming 47-779 Ising Model 1 William Larimer Mellon, Founder Agenda o

Quantum Integer Programming 47-779 Ising Model 1 William Larimer Mellon, Founder Agenda o

Quantum Integer Programming 47-779 Ising Model 1 William Larimer Mellon, Founder Agenda o Ising Model: Background, Physics o Ising Model: Solutions o Ising Model and Combinatorial Optimization o Ising Model and Integer Programming o Solving

669 views • 19 slides
Random triangulations coupled with an Ising model Laurent M enard (Paris Nanterre) joint work

Random triangulations coupled with an Ising model Laurent M enard (Paris Nanterre) joint work

Random triangulations coupled with an Ising model Laurent M enard (Paris Nanterre) joint work with Marie Albenque and Gilles Schaeffer (CNRS and LIX) Bordeaux, November 2018 Outline 1. Introduction: 2DQG and planar maps 2. Local weak

1.1k views • 88 slides
Height representation of XOR-Ising loops via bipartite dimers C edric Boutillier (UPMC) B

Height representation of XOR-Ising loops via bipartite dimers C edric Boutillier (UPMC) B

Height representation of XOR-Ising loops via bipartite dimers C edric Boutillier (UPMC) B eatrice de Tili` ere (UPMC) LAGA Universit e Paris Nord March 11, 2015 The Ising model and the XOR-Ising model The Ising model Let G =

797 views • 43 slides
The critical Z -invariant Ising model via dimers B eatrice de Tili` ere University of

The critical Z -invariant Ising model via dimers B eatrice de Tili` ere University of

The critical Z -invariant Ising model via dimers B eatrice de Tili` ere University of Neuch atel joint work with C edric Boutillier, University Paris 6 Ascona, 27 May 2010 The 2 -dimensional Ising model Planar graph, G = ( V ( G )

600 views • 24 slides
1st PRIMA Congress Conformal invariance and universality in the 2D Ising model Stanislav Smirnov

1st PRIMA Congress Conformal invariance and universality in the 2D Ising model Stanislav Smirnov

1st PRIMA Congress Conformal invariance and universality in the 2D Ising model Stanislav Smirnov July 6, 2009 Archetypical example of a phase transition: 2D Ising model Configurations of + and spins (red and blue squares) Prob( config )

669 views • 31 slides
SLE and conformal invariance for critical Ising model Stanislav Smirnov jointly with Dmitry

SLE and conformal invariance for critical Ising model Stanislav Smirnov jointly with Dmitry

3rd LA PIETRA WEEK IN PROBABILITY Stochastic Models in Physics Firenze, June 23-27, 2008 SLE and conformal invariance for critical Ising model Stanislav Smirnov jointly with Dmitry Chelkak 1D Ising model 0 1

355 views • 20 slides
Crossing probabilities in the critical 2D Ising model Dmitry Chelkak (PDMI Steklov, St.Petersburg)

Crossing probabilities in the critical 2D Ising model Dmitry Chelkak (PDMI Steklov, St.Petersburg)

Crossing probabilities in the critical 2D Ising model Dmitry Chelkak (PDMI Steklov, St.Petersburg) joint work with Stanislav Smirnov (Geneva) arXiv:0910.2045 : Universality in the 2D Ising model and conformal invariance of fermionic

667 views • 42 slides
Planar Ising model at criticality: state-of-the-art and perspectives Dmitry Chelkak, ENS,

Planar Ising model at criticality: state-of-the-art and perspectives Dmitry Chelkak, ENS,

Planar Ising model at criticality: state-of-the-art and perspectives Dmitry Chelkak, ENS, Paris [ &amp; PDMI, St. Petersburg (on leave) ] ICM, Rio de Janeiro, August 4, 2018 Planar Ising model at criticality: outline Combinatorics

791 views • 37 slides
First order phase transition for the Random Cluster model with q &gt; 4 Ioan Manolescu joint work

First order phase transition for the Random Cluster model with q &gt; 4 Ioan Manolescu joint work

First order phase transition for the Random Cluster model with q &gt; 4 Ioan Manolescu joint work with: Hugo Duminil-Copin, Maxime Gagnebin, Matan Harel, Vincent Tassion University of Fribourg 14th February 2017 Diablerets Ioan Manolescu

830 views • 81 slides
Monte Carlo simulation of the 2D Ising model - tutorial - Zoltn Nda Babe-Bolyai University

Monte Carlo simulation of the 2D Ising model - tutorial - Zoltn Nda Babe-Bolyai University

Monte Carlo simulation of the 2D Ising model - tutorial - Zoltn Nda Babe-Bolyai University Department of Theoretical and Computational Physics . A basic Metropolis Algorithm for simulating the 2D and 3D Ising model on square lattice

462 views • 8 slides
Improved Clustering Algorithms for the Random Cluster Graph Model Ron Shamir Dekel Tsur Tel

Improved Clustering Algorithms for the Random Cluster Graph Model Ron Shamir Dekel Tsur Tel

Improved Clustering Algorithms for the Random Cluster Graph Model Ron Shamir Dekel Tsur Tel Aviv University 1/18 The Clustering Problem Input: A graph G . (edges in G represent similarity between the vertices) Output: A partition of the

485 views • 34 slides
Ising Model and Percolation Problem Shahin Rouhani Physics Department Sharif University of

Ising Model and Percolation Problem Shahin Rouhani Physics Department Sharif University of

Statistical Mechanics of Two Dimensional Critical Curves Ising Model and Percolation Problem Shahin Rouhani Physics Department Sharif University of Technology Tehran, Iran. Scale invariance in the critical Ising model Let us illustrate

394 views • 27 slides
Convergence of correlations in the 2D Ising model: primary fields [ and the stress-energy tensor

Convergence of correlations in the 2D Ising model: primary fields [ and the stress-energy tensor

Convergence of correlations in the 2D Ising model: primary fields [ and the stress-energy tensor ] Dmitry Chelkak [ ENS, Paris &amp; Steklov Institute, St. Petersburg ] [ Sample of a critical 2D Ising configuration (with two disorders), c

1.24k views • 54 slides
The Prokofiev-Svistunov-Ising process is rapidly mixing Tim Garoni School of Mathematical

The Prokofiev-Svistunov-Ising process is rapidly mixing Tim Garoni School of Mathematical

Introduction Main Theorem Proof Discussion The Prokofiev-Svistunov-Ising process is rapidly mixing Tim Garoni School of Mathematical Sciences Monash University Introduction Main Theorem Proof Discussion Collaborators Andrea

1.84k views • 147 slides
DECISION MAKING IN A CONDOMINIUM AN ISING-LIKE SOCIOPHYSICAL SYSTEM Mrta Jvor Tams

DECISION MAKING IN A CONDOMINIUM AN ISING-LIKE SOCIOPHYSICAL SYSTEM Mrta Jvor Tams

DECISION MAKING IN A CONDOMINIUM AN ISING-LIKE SOCIOPHYSICAL SYSTEM Mrta Jvor Tams Geszti Roland Etvs University IN MEMORIAM GEORGE MARX (1927 2002) 2 Purpose to show: how we can use a simple model of statistical

556 views • 55 slides
Ferromagnets and superconductors Kay Kirkpatrick, UIUC Ferromagnet and superconductor models:

Ferromagnets and superconductors Kay Kirkpatrick, UIUC Ferromagnet and superconductor models:

Ferromagnets and superconductors Kay Kirkpatrick, UIUC Ferromagnet and superconductor models: Phase transitions and asymptotics Kay Kirkpatrick, Urbana-Champaign October 2012 Ferromagnet and superconductor models: Phase transitions and

1k views • 58 slides
Holographic encoding of universality in corner spectra National Center for Theoretical Sciences

Holographic encoding of universality in corner spectra National Center for Theoretical Sciences

Holographic encoding of universality in corner spectra National Center for Theoretical Sciences (NCTS),Taiwan Ching-Yu Huang ( ) work with Romn Ors (Donostia International Physics Center) Tzu-Chieh Wei (Stony Brook university)

460 views • 44 slides
Bootstrapping the 3D Ising Model David Simmons-Duffin IAS Strings 2014 with S. El-Showk, M.

Bootstrapping the 3D Ising Model David Simmons-Duffin IAS Strings 2014 with S. El-Showk, M.

Bootstrapping the 3D Ising Model David Simmons-Duffin IAS Strings 2014 with S. El-Showk, M. Paulos, F. Kos, D. Poland, S. Rychkov, A. Vichi The Conformal Bootstrap Polyakov 70: classify/solve CFTs using: conformal symmetry unitarity

445 views • 20 slides
High-dimensional Ising model and Monte Carlo methods Wojciech Rejchel Nicolaus Copernicus

High-dimensional Ising model and Monte Carlo methods Wojciech Rejchel Nicolaus Copernicus

Introduction Model selection consistency Experiments References High-dimensional Ising model and Monte Carlo methods Wojciech Rejchel Nicolaus Copernicus University in Toru Joint work with Baej Miasojedow Wojciech Rejchel

1.01k views • 84 slides
Inference and Sampling of K 33 -free Ising Models Valerii Likhosherstov 1 , Yury Maximov 1,2 ,

Inference and Sampling of K 33 -free Ising Models Valerii Likhosherstov 1 , Yury Maximov 1,2 ,

Inference and Sampling of K 33 -free Ising Models Valerii Likhosherstov 1 , Yury Maximov 1,2 , Michael Chertkov 1,2,3 1 Skolkovo Institute of Science and Technology, Moscow, Russia 2 Theoretical Division and Center for Nonlinear Studies, Los Alamos

797 views • 7 slides
CSci 8980: Advanced Topics in Graphical Models Variational Inference Instructor: Arindam Banerjee

CSci 8980: Advanced Topics in Graphical Models Variational Inference Instructor: Arindam Banerjee

Graphical Models Exponential Families Variational Methods Mean Field Approximation CSci 8980: Advanced Topics in Graphical Models Variational Inference Instructor: Arindam Banerjee October 17, 2007 Graphical Models Exponential Families

1.24k views • 106 slides

More recommend