A New Hybrid Algorithm for Noisy Optimization (Learning Automata + Genetic Algorithm)

The learning automata operate in unknown random environments and progressively improve their performance via a learning process. The learning automata are very useful for optimization of multi-modal functions when the function is unknown and only noise-corrupted evaluations are available. In this paper we propose a new hybrid algorithm for noisy optimization. This model is obtained by combining continuous action set learning automata and evolutionary algorithm. In order to show the performance of the proposed hybrid algorithm it is tested on a function optimization problem. The results of experimentation show the superiority of the proposed model over the two existing learning automata based algorithm for optimization.