一种用于神经网络汉字识别系统的自组织聚类方法

A New Clustering Method for Chinese Character Recognition System Using Neural Networks

本文提出一种用于大类别模式识别系统的改进的自组织聚类方法．采用这种方法把３７５５种多体印刷汉字聚类为几千个子集．聚类正确率优于９８％．

This paper proposes a new clustering method for Chinese character recognition system using ANN. By use of this clustering technique,the large Chinese character set of 3755 characters can be divided into many small subsets with accuracy higher than 98%.