Fuzzy-valued evolution strategy for evolving neural networks with fuzzy weights and biases
Hidehiko Okada, Akira Yamashita, Takashi Matsuse, Tetsuya Wada · 2012
In this paper, we propose an extension of evolution strategy (ES) for evolving fuzzy-valued neural networks (FNNs). In the proposed ES, values in the genotypes are not real numbers but fuzzy values. We apply our fuzzy-valued ES (FES) to the approximate modeling of fuzzy functions with FNNs. Experimental results showed that an FNN trained by our FES could approximate a hidden test function to a certain extent, despite t that the learning was not supervised.