Non-Stationary Stochastic Global Optimization Algorithms

Studying the theoretical properties of optimization algorithms such as genetic and evolutionary strategies allows us to determine when they are suitable for solving a particular type problem. Such study consists three main steps. The first step is considering Stochastic Global Optimization Algorithms (SGoals ), i.e., iterative algorithm that applies stochastic operations set candidate solutions. second define formal characterization process in terms measure theory some stationary Markov kernels (defined transition probabilities do not change over time). third characterize non-stationary SGoals, SGoals having with may time. In this paper, we develop study. First, generalize sufficient conditions convergence from processes. Second, introduce necessary arithmetic between measurable functions. Third, selection recombination schemes. Finally, formalize simulated annealing using systematic approach.