基于Bayes分类的PAC-Bayesian定理证明  

Proof of PAC-Bayesian Theorem Based on Bayes Classifier

作　　者：刘妙妙[1] 赵联文[1] 宋海龙[1] 姜英[1] 

机构地区：[1]西南交通大学数学学院,四川成都610031

出　　处：《吉首大学学报（自然科学版）》2013年第6期19-21,70,共4页Journal of Jishou University(Natural Sciences Edition)

基　　金：中央高校基本科研业务费专项资金资助(SWJTU11CX155)

摘　　要：采用类似基于Gibbs分类的PAC-Bayesian定理的证明方法,证明基于Bayes分类的PAC-Bayesian定理.对于PAC-Bayesian定理的证明采用Bayes分类,可以方便有效地运用到统计问题中来解决相关问题.Method of proving PAC-bayesian theorem based on Gibbs classification is used to the proof based on Bayes classification. Through the Bayes classification,the theorem can be applied to the statisti- cal problems conveniently and effectively to solve related problems.

关 键 词：Bayesian方法 BAYES分类 PAC-Bayesian 

分 类 号：O212.8[理学—概率论与数理统计]