基于最小错误率的贝叶斯决策在手写英文字母分类识别中的应用  

Bayes Decision for Minimum Errors Applied in Recognition of Handwritten English Letters

作　　者：荆钟[1] 何明[1] 

机构地区：[1]辽宁工业大学电子与信息工程学院,辽宁锦州121001

出　　处：《辽宁工业大学学报（自然科学版）》2009年第2期98-100,107,共4页Journal of Liaoning University of Technology(Natural Science Edition)

摘　　要：统计决策理论是处理模式识别问题的基本理论之一,而贝叶斯决策理论方法又是统计模式识别中的一个基本方法,它可以有效地对大量数据进行分析,并生成相应的分类器,对于数据的分类识别有着重大的意义。把最小错误率的贝叶斯方法运用到手写英文字母的识别中,提高了分类的准确性和有效性。The statistical decision-making theory was one of the basic theories for treating the problem on pattern recognition, however, the method of Bayesian Decision-making Theory was the basic one in pattern recognition, facing the massive data, which can be used to make effective analysis, and produce corresponding sorters, thus possessing important significance for the classification and recognition of the data. Bayes decision for minimum errors applied in recognition of handwritten English letters improved the exactitude and effiectiveness in classification.

关 键 词：贝叶斯理论 模式识别 统计决策 手写英文字母 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]