An accelerated particle swarm optimized back propagation algorithm

Recently,accelerated particle swarm optimization (apso) derived from particle swarm optimization (pso) algorithm’s principle is becoming a very popular method in solving many hard optimization problems particularly the inherent weight problem in back propagation (bp). therefore,this paper proposed an accelerated particle swarm optimized back propagation neural network (apso-bp) algorithm in order to overcome the problems faced in bp algorithm. by using apso to optimize the weights at each iterations of bp algorithm,the proposed apso-bp is able to increase the convergence speed and avoids local minima. the simulation results demonstrates that the proposed algorithm outperforms the traditional bp method and achieves the objectives of this research,which contributes to artificial intelligence field. © 2015 penerbit utm press. all rights reserved.