A Simplified Learning Algorithm of Incremental Bayesian

Author

Hua, Chen ; Xiao-Gang, Zhang ; Jing, Zhang ; Li-hua, Ding

Author_Institution

Sch. of Comput. & Commun., Hunan Univ., Changsha, China

Volume

5

fYear

2009

fDate

March 31 2009-April 2 2009

Firstpage

126

Lastpage

128

Abstract

A proportion factor is constructed though the Maximum Aposteriori Probability of examples in test data to select the training examples in incremental learning process. Instead of complex normal classify loss expression, the proportion factor lambda is used to estimate the classify loss to improve classification efficiency. The final experiment shows that this algorithm is feasible, and more accurate than simple Bayesian classifier. The computing time is highly reduced on the optimal selection of examples in incremental learning process.

Keywords

belief networks; learning (artificial intelligence); pattern classification; probability; Bayesian classifier; Probability; complex normal classify loss expression; incremental Bayesian; optimal selection; simplified learning algorithm; Algorithm design and analysis; Bayesian methods; Computer science; Data engineering; Data mining; Machine learning; Maximum a posteriori estimation; Probability distribution; Testing; Training data; Bayesian classifier; incremental learning; simplified algorithm;

fLanguage

English

Publisher

ieee

Conference_Titel

Computer Science and Information Engineering, 2009 WRI World Congress on

Conference_Location

Los Angeles, CA

Print_ISBN

978-0-7695-3507-4

Type

conf

DOI

10.1109/CSIE.2009.305

Filename

5170510