Simulated Annealing in the Variable Landscape
Manabu Hasegawa, Chang Ju Kim · Proceedings of the 12th Asia Pacific Physics Conference (APPC12) · 2014
An experimental analysis is conducted to test whether the appropriate introduction of the smoothness-temperature schedule enhances the optimizing ability of the MASSS method, the combination of the Metropolis algorithm (MA) and the search-space smoothing (SSS) method.The test is performed on two types of random traveling salesman problems.The results show that the optimization performance of the MA is substantially improved by a single smoothing alone and slightly more by a single smoothing with cooling and by a de-smoothing process with heating.The performance is compared to that of the parallel tempering method and a clear advantage of the idea of smoothing is observed depending on the problem.