Mutually beneficial relationship in optimization between search-space smoothing and stochastic search

The effectiveness of the Metropolis algorithm (MA) (constant-temperature simulated annealing) in optimization by the method of search-space smoothing (SSS) (potential smoothing) is studied on two types of random traveling salesman problems. The optimization mechanism of this hybrid approach (MASSS) is investigated by analyzing the exploration dynamics observed in the rugged landscape of the cost function (energy surface). The results show that the MA can be successfully utilized as a local search algorithm in the SSS approach. It is also clarified that the optimization characteristics of these two constituent methods are improved in a mutually beneficial manner in the MASSS run. Specifically, the relaxation dynamics generated by employing the MA work effectively even in a smoothed landscape and more advantage is taken of the guiding function proposed in the idea of SSS; this mechanism operates in an adaptive manner in the de-smoothing process and therefore the MASSS method maintains its optimization function over a wider temperature range than the MA.

[1]  Giorgio Parisi,et al.  On the approach to the equilibrium and the equilibrium properties of a glass-forming model , 1998 .

[2]  Elmar Schömer,et al.  Packing a multidisperse system of hard disks in a circular environment. , 2009, Physical review. E, Statistical, nonlinear, and soft matter physics.

[3]  Bruce L. Golden,et al.  A computational study of smoothing heuristics for the traveling salesman problem , 2000, Eur. J. Oper. Res..

[4]  Thomas A. Weber,et al.  Hidden structure in liquids , 1982 .

[5]  Jay W. Ponder,et al.  Exploring the similarities between potential smoothing and simulated annealing , 2000 .

[6]  Peter Salamon,et al.  Facts, Conjectures, and Improvements for Simulated Annealing , 1987 .

[7]  Jun Gu,et al.  Efficient Local Search With Search Space Smoothing: A Case Study of the Traveling Salesman Problem (TSP) , 1994, IEEE Trans. Syst. Man Cybern. Syst..

[8]  K. Hukushima,et al.  Exchange Monte Carlo Method and Application to Spin Glass Simulations , 1995, cond-mat/9512035.

[9]  F. Stillinger,et al.  Packing Structures and Transitions in Liquids and Solids , 1984, Science.

[10]  M. Hasegawa,et al.  Positive role of glassy dynamics in finite-time optimization by threshold algorithms , 2011, Comput. Phys. Commun..

[11]  M Hasegawa.  Evaluation of the physical annealing strategy for simulated annealing: a function-based analysis in the landscape paradigm. , 2012, Physical review. E, Statistical, nonlinear, and soft matter physics.

[12]  N. Metropolis,et al.  Equation of State Calculations by Fast Computing Machines , 1953, Resonance.

[13]  Manabu Hasegawa.  Unidirectional Search Dynamics in Optimization by Metropolis Algorithm , 2005 .

[14]  J. M. Singer,et al.  Search-space smoothing for combinatorial optimization problems , 1997 .

[15]  Srikanth Sastry,et al.  Signatures of distinct dynamical regimes in the energy landscape of a glass-forming liquid , 1998, Nature.

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

[17]  V. Cerný.  Thermodynamical approach to the traveling salesman problem: An efficient simulation algorithm , 1985 .

[18]  G. Parisi,et al.  Simulated tempering: a new Monte Carlo scheme , 1992, hep-lat/9205018.

[19]  H. Scheraga,et al.  On the multiple-minima problem in the conformational analysis of molecules: deformation of the potential energy hypersurface by the diffusion equation method , 1989 .

[20]  Thomas A. Weber,et al.  Dynamics of structural transitions in liquids , 1983 .

[21]  M Hasegawa,et al.  Verification and rectification of the physical analogy of simulated annealing for the solution of the traveling salesman problem. , 2011, Physical review. E, Statistical, nonlinear, and soft matter physics.

[22]  I. Morgenstern,et al.  Influence of rounding errors on the quality of heuristic optimization algorithms , 2011 .

[23]  Eugene L. Lawler,et al.  The Traveling Salesman Problem: A Guided Tour of Combinatorial Optimization , 1985 .

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