论文部分内容阅读
神经网络能够正确记忆的样本矢量数与其神经元数目的比称为网络的存储容量,它是神经网络的一个重要参数。对Hopfield神经网络的存储容量的研究,统计物理方法分析的结果约为0.138。研究表明,当网络的存储容量为
The ratio of the number of sample vectors that can be correctly memorized to the number of neurons in the neural network is called the storage capacity of the network. It is an important parameter of the neural network. The research on the storage capacity of Hopfield neural network, the result of statistical physical method analysis is about 0.138. Research shows that when the network storage capacity is