基于差分进化的BP网络学习算法

A BP neural network learning algorithm based on differential evolution

提出了一种基于差分进化的BP网络学习算法,该算法是一种全局随机优化算法.利用差分进化算法的全局寻优能力,可以很好的训练BP网络的权值和阈值.将所提出的算法与BP算法作对比实验,结果表明,所提出的算法相对于BP算法在分类准确度上有较大的提高,而且具有良好的收敛性和泛化能力.

A BP neural per and the algorithm tion, the algorithm is works learning algorithm based on differential evolution is proposed in this pastochastic global optimization technique. Using the ability of global optimizaone best in training the weight values and threshold values of BP networks. Comparing experiments between the proposed algorithm and BP algorithm, the results show that the proposed algorithm does not only superior to BP algorithm in classification precision , bus also has fast convergence speed and generalization ability.