Performance Limitations of Flat Histogram Methods and Optimality of Wang-Landau Sampling

Abstract

We determine the optimal scaling of local-update flat-histogram methods with system size by using a perfect flat-histogram scheme based on the exact density of states of 2D Ising models.The typical tunneling time needed to sample the entire bandwidth does not scale with the number of spins N as the minimal N^2 of an unbiased random walk in energy space. While the scaling is power law for the ferromagnetic and fully frustrated Ising model, for the +/- J nearest-neighbor spin glass the distribution of tunneling times is governed by a fat-tailed Frechet extremal value distribution that obeys exponential scaling. We find that the Wang-Landau algorithm shows the same scaling as the perfect scheme and is thus optimal.

5 pages, 6 figures

Cited by in corpus (34)

• Continuous-time Monte Carlo methods for quantum impurity models
• Feedback-optimized parallel tempering Monte Carlo
• Optimized parallel tempering simulations of proteins
• Optimizing the ensemble for equilibration in broad-histogram Monte Carlo simulations
• Wang-Landau Algorithm: a Theoretical Analysis of the Saturation of the Error
• Generalized Ensemble and Tempering Simulations: A Unified View
• Optimal Modification Factor and Convergence of the Wang-Landau Algorithm
• Football fever: goal distributions and non-Gaussian statistics
• Entropic sampling via Wang-Landau random walks in dominant energy subspaces
• First-order transition features of the 3D bimodal random-field Ising model
• Parallel multicanonical study of the three-dimensional Blume-Capel model
• Quenched bond randomness in marginal and non-marginal Ising spin models in 2D
• Dynamics of the Wang-Landau algorithm and complexity of rare events for the three-dimensional bimodal Ising spin glass
• Optimized ensemble Monte Carlo simulations of dense Lennard-Jones fluids
• Overcoming the critical slowing down of flat-histogram Monte Carlo simulations: Cluster updates and optimized broad-histogram ensembles
• Universality aspects of the 2d random-bond Ising and 3d Blume-Capel models
• First-order transition features of the triangular Ising model with nearest- and next-nearest-neighbor antiferromagnetic interactions
• Manifold Learning in Atomistic Simulations: A Conceptual Review
• Lower bounds for Ramsey numbers as a statistical physics problem
• Optimized multicanonical simulations: a new proposal based on classical fluctuation theory
• Optimized Broad-Histogram Ensembles for the Simulation of Quantum Systems
• Optimized broad-histogram simulations for strong first-order phase transitions: Droplet transitions in the large-Q Potts model
• Multicanonical Sampling of the Space of States of H(2,n)-Vector Models
• Universal features and tail analysis of the order-parameter distribution of the two-dimensional Ising model: An entropic sampling Monte Carlo study
• Efficiency of Monte Carlo Sampling in Chaotic Systems
• Analytical study of tunneling times in flat histogram Monte Carlo
• Intrinsic convergence properties of entropic sampling algorithms
• Energy gap of the bimodal two-dimensional Ising spin glass
• Perturbation method to calculate the density of states
• Applying the Wang-Landau Algorithm to Lattice Gauge Theory
• Typical versus average helicity modulus in the three-dimensional gauge glass: Understanding the vortex glass phase
• Avoiding critical slowdown in models with SALR interactions
• Quantum Annealing Algorithms for Estimating Ising Partition Functions
• Projected single-spin flip dynamics in the Ising Model