Metropolis, Simulated Annealing, and Iterated Energy Transformation Algorithms: Theory and Experiments

Abstract In this paper, we compare from the theoretical and experimental points of view three stochastic optimization algorithms: the Metropolis, simulated annealing, and iterated energy transformation algorithms. We give the optimal exponents for the concentration of the marginal distribution of the final state of these algorithms around the global minima of the virtual energy function. Experiments are performed on an N.P. complete benchmark which tries to retain the main aspects of scheduling problems. They lead to the same qualitative ranking of algorithms as the theory does.

[1]  A. Trouvé.  Cycle Decompositions and Simulated Annealing , 1996 .

[2]  L. Miclo.  Sur les problèmes de sortie discrets inhomogènes , 1996 .

[3]  O. Catoni.  The energy transformation method for the Metropolis algorithm compared with Simulated Annealing , 1998 .

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

[5]  T. Chiang,et al.  A Limit Theorem for a Class of Inhomogeneous Markov Processes , 1989 .

[6]  O. Catoni,et al.  Piecewise constant triangular cooling schedules for generalized simulated annealing algorithms , 1998 .

[7]  L. Miclo.  Evolution de l'énergie libre, applications à l'étude de la convergence des algorithmes du recuit simulé , 1991 .

[8]  D. Stroock,et al.  Simulated annealing via Sobolev inequalities , 1988 .

[9]  J. Deuschel,et al.  $L^2$ Convergence of Time Nonhomogeneous Markov Processes: I. Spectral Estimates , 1994 .

[10]  A. Trouvé.  Parallélisation massive du recuit simulé , 1993 .

[11]  O. Catoni,et al.  Exponential triangular cooling schedules for simulated annealing algorithms : A case study , 1992 .

[12]  O. Catoni,et al.  The exit path of a Markov chain with rare transitions , 1997 .

[13]  M. Freidlin,et al.  Random Perturbations of Dynamical Systems , 1984 .

[14]  O. Catoni.  Solving Scheduling Problems by Simulated Annealing , 1998 .

[15]  C. Hwang,et al.  Singular perturbed Markov chains and exact behaviors of simulated annealing processes , 1992 .

[16]  Donald Geman,et al.  Stochastic relaxation, Gibbs distributions, and the Bayesian restoration of images , 1984 .

[17]  John N. Tsitsiklis,et al.  Markov Chains with Rare Transitions and Simulated Annealing , 1989, Math. Oper. Res..

[18]  M. Laurent.  Sur les temps d'occupations des processus de markov finis inhomogènes à basse température , 1998 .

[19]  A. Trouvé.  Rough Large Deviation Estimates for the Optimal Convergence Speed Exponent of Generalized Simulated , 1994 .

[20]  Robert Azencott,et al.  Simulated annealing : parallelization techniques , 1992 .

[21]  David S. Johnson,et al.  Computers and Intractability: A Guide to the Theory of NP-Completeness , 1978 .

[22]  O. Catoni.  Rough Large Deviation Estimates for Simulated Annealing: Application to Exponential Schedules , 1992 .

[23]  P. Diaconis,et al.  Geometric Bounds for Eigenvalues of Markov Chains , 1991 .