Measure-theoretic evolutionary annealing

Abstract

Abstract-There is a deep connection between simulated annealing and genetic algorithms with proportional selection. Evolutionary annealing is a novel evolutionary algorithm that makes this connection explicit, resulting in an evolutionary optimization method that can be viewed either as simulated annealing with improved sampling or as a non-Markovian selection mechanism for genetic algorithms with selection over all prior populations.A martingale-based analysis shows that evolutionary annealing is asymptotically convergent and this analysis leads to heuristics for setting learning parameters to optimize the convergence rate. In this work and in parallel work evolutionary annealing is shown to converge faster than other evolutionary algorithms on several benchmark problems, establishing a promising foundation for future theoretical and experimental research into algorithms based on evolutionary annealing.

Determining the significance and relative importance of parameters of a simulated quenching algorithm using statistical tools

“…SA is a generic probabilistic metaheuristic method for solving global optimization problems. Subject to conditions on the cooling schedule, simulated annealing can be shown to converge asymptotically to the global optima of the fitness function [16,5,62,63].…”

“…In practice, simulated annealing has been used effectively in several science and engineering problems [64,65,66]. However, its sensitivity to the proposal distribution and the cooling schedule means that it is not a good fit for all optimization problems [62,63]. In a SA algorithm, a cost function (objective function f ) to be minimized is defined.…”

“…SA strength relies in its ability to statistically deliver the system to a global optimum, although in some cases the convergence could only be obtained after an infinite number of iterations (as stated in [62,63]). However, due to its sensitivity to the proposal distribution and the cooling schedule, other algorithms (i.e.…”