一种大类别数分类的神经网络方法

神经网络是一种普遍使用的分类方法。当类别数目较大时，神经网络结构复杂、训练时间激增、分类性能下降。针对这些问题，基于Ｎ分类问题的两种类方法和树型分类器结构，对两分类子网络集进行排序，中给出了一种大类别分类的神经网络阵一结构和快速搜索方法并重点分析了网络阵列的分类性能。理论分析表明，使用网络阵列方法可降低平均分类错误率。该方法还使得网络结构简单灵活，易于扩充，网络的训练时间缩短，仿真实验表明，该方

A NEURAL NETWORK METHOD OF CLASSIFICATION FOR LARGE NUMBER OF CATALOGS

Neural network has been used for pattern recognition popularly. The training time of neural network for N catalogs classification increases exponentially with N, so it is difficult to deal with large number of catalogs by normal neural networks. Based on binary partition method and decision tree, a neural network array for classifying large number of catalogs is proposed in this paper. Each element in the array is a simple neural network which only processes 2 catalog classification. Thus the architecture of network array is flexible and expansible, and the training time is reduced largely. It is proved that the mean error of classifying N catalogs with neural network array is smaller than that with a normal neural network. Experiment shows that this method can classify large number of catalogs well.