论文信息 - Blocked stochastic sampling versus Estimation of Distribution Algorithms

Blocked stochastic sampling versus Estimation of Distribution Algorithms

The Boltzmann distribution is a good candidate for a search distribution for optimization problems. We compare two methods to approximate the Boltzmann distribution - Estimation of Distribution Algorithms (EDA) and Markov Chain Monte Carlo methods (MCMC). It turns out that in the space of binary functions even blocked MCMC methods outperform EDA on a small class of problems only. In these cases a temperature of T = 0 performed best.

Heinz Mühlenbein | Roberto Santana | H. Mühlenbein | Roberto Santana

[1] Heinz Mühlenbein,et al. Schemata, Distributions and Graphical Models in Evolutionary Optimization , 1999, J. Heuristics.

[2] L. Kallel,et al. Theoretical Aspects of Evolutionary Computing , 2001, Natural Computing Series.

[3] Thomas Bäck,et al. An Overview of Evolutionary Algorithms for Parameter Optimization , 1993, Evolutionary Computation.

[4] John M. Olin. On MCMC sampling in hierarchical longitudinal models , 1999 .

[5] H. Muhlenbein,et al. Size of neighborhood more important than temperature for stochastic local search , 2000, Proceedings of the 2000 Congress on Evolutionary Computation. CEC00 (Cat. No.00TH8512).

[6] Jun S. Liu,et al. Covariance structure of the Gibbs sampler with applications to the comparisons of estimators and augmentation schemes , 1994 .

[7] J. Propp,et al. Exact sampling with coupled Markov chains and applications to statistical mechanics , 1996 .

[8] Donald Geman,et al. Stochastic Relaxation, Gibbs Distributions, and the Bayesian Restoration of Images , 1984, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[9] C. D. Gelatt,et al. Optimization by Simulated Annealing , 1983, Science.