摘要
为防止交叉后优秀基因段的丢失,在随机非一致线性交叉的基础上,设计了一种与个体适应度相关的线性交叉方案。构造了一种使交叉率与变异率随进化过程自适应调整的方法,有效抑制了遗传算法的早熟收敛。然后,针对函数逼近问题用改进后的遗传算法去优化前馈神经网络的结构,降低了神经网络训练陷入局部最优的可能性,提高了网络的泛化能力。
To prevent the loss of good gene section after crossover,a linear crossover program related to individual fitness is designed on the ground of random non-linear crossover. By adjusting crossover rate and mutation rate adaptively, the premature convergence of the algorithm is effectively suppressed. Then, directing at the function approximation problem, the improved genetic algorithm is used to optimize the structure of feed-forward neural network. This method reduces the possibility of getting local optimum in neural network training and enhances the generalization capability of network
出处
《系统仿真技术》
2013年第2期170-174,共5页
System Simulation Technology
关键词
遗传算法
神经网络
结构优化
函数逼近
genetic algorithm
neural network
structure optimization
function approximation