基于加强贝叶斯分类的手写体数字识别  

Handwriting Digit Recognition Based on Enhanced Bayes Classification

作　　者：罗成[1] 孙越恒[1] 

机构地区：[1]天津大学计算机科学与技术学院,天津300072

出　　处：《微处理机》2009年第3期77-79,共3页Microprocessors

摘　　要：提出了一种手写体数字识别的新方案。在方案中,设计了一种加强贝叶斯分类器,该分类器可以对训练集进行反馈式学习。经过一定强度的训练,能够对测试过程加以监督和修正。通过对实际样本的测试证实该方法的有效性。This paper proposes a new method of handwriting digit recognition. In this new method, the paper designs a new type of enhanced Bayes classifier, which could process study - under - feed - back on training data. After some certain intensity of training, the classifier can implement supervision and correction to the process of testing. The validity of this method is authenticated when completed the experiments on practical data.

关 键 词：手写体识别 贝叶斯分类 反馈式学习 

分 类 号：TP391.43[自动化与计算机技术—计算机应用技术]