【摘要】 隐层神经元冗余是提高神经网络容错性的一个有效的方法 ,在神经网络分类器的容错设计中 ,这一方法得到了良好的效果 ,对单故障可以做到完全容错 .但是这一应用仅仅只能应用于输出层为硬限幅函数的前向网络 ,并且只证明了对网络中单故障有效 .在实际应用中 ,网络中的各个节点和权值的故障往往是普遍存在的 ,因此本文提出了一种隐层冗余结构 ,对普遍故障存在下隐层神经元冗余容错方法做以评估 ,得出的结论是应用这种隐层神经元冗余结构可以减小网络的全局故障率 ;并提出了针对一般前向神经网络的实用的隐层神经元容错方法 ,这种方法可以有效地提高网络在普遍故障下的容错能力 .更多还原

【Abstract】 Redundancy on hidden layer neurons has been proven useful in the fault tolerance of neural networks.This approach has been applied successfully in the fault tolerance design of classification neural networks,thus the complete single fault tolerance can be gained.But this approach can only be applied to the feed forward networks which has hard limit activation functions in output layer.And it is proved that this approach is valid only to single fault.There are universal faults of neurons and weights in actual applications,so we evaluated this approach under universal faults in feed forward networks.We proved that the global fault rate is reduced though the redundancy on hidden layer neurons.And we presented a practical and valid method of redundancy of hidden layer neurons to gain fault tolerance.更多还原