DNA Algorithm for Training Fuzzy Neural Networks
Xikui Liu · Journal of Chinese Computer Systems · 2006
A DNA genetic algorithm is proposed based on DNA computing and genetic algorithms.The structure of the DNA genetic algorithms is presented,and the operators of the proposed algorithm are discussed.Compared with the grads algorithm and the genetic algorithms,the DNA genetic algorithm owns more accurate precision and faster convergent speed when it is used to train FNN.The algorithm obtains the performance of global convergence which avoids the inherent convergence when the grids algorithm is adopted to train FNN.Also the proposed algorithm accelerates FNN's training and can be used on-line.