利用遗传算法改善前馈神经网络容错性

IMPROVING FAULT-TOLERANCE OF FEEDFORWARD NEURAL NETWORKS WITH GENETIC ALGORITHMS

针对前馈神经网络的断路故障 ,将网络容错性的改善转化为一个最小优化问题 ,并通过遗传算法进化求解来获取容错性好、泛化能力强的网络 .该方法不需给网络增加额外冗余 ,也不需修改网络训练算法 ,较好地保持了网络结构、训练算法与容错处理的独立性 .实验表明 。

The problem of improving the fault-tolerance of feedforward neural networks that suffers open fault is transferred to a minimum optimization problem, and a genetic algorithm is used to evolve networks that have good fault-tolerance and strong generalization ability. Since neither extra network redundancy is introduced nor training algorithm is modified, the independence of the network architecture, training algorithm, and fault-tolerant process are well kept. Experiments show that this method achieves good results in two benchmark tests.