Adaptation of neural agent in dynamic environment: hybrid system of genetic algorithm and neural network
Takashi Iba, Yoshiyasu Takefuji · 2002
This study proposes an adaptive agent as a hybrid of genetic algorithm and neural network, and to clarify the effectiveness of the combination of two mechanisms in the dynamic environment. Evolution and learning can be explained as the mechanism of searching a solution in the enormous possibilities at the population level and individual level, respectively. There are two ways of combination of genetic algorithm and neural network: Darwinian and Lamarckian frameworks. In the Lamarckian framework the acquired traits during the lifetime can be passed on to the offspring directly, while in the Darwinian framework, these cannot be passed on. We propose a "neural agent" whose initial weights of their neural networks are determined by their genome data, as a simple model of the hybrid system of genetic algorithm and neural network. We examine which framework is better in the dynamic system. The result of our simulation shows that the Darwinian framework is better than Lamarckian.