Bayesian classifier based on discretized continuous feature space

Author

Zhou, Dequan ; Wu, Liguang ; Liu, GuoSui

Author_Institution

Air Force No.1 Inst. of Aeronaut., Xinyang City, China

Volume

2

fYear

1998

fDate

1998

Firstpage

1225

Abstract

Bayesian decision theory is widely used in pattern recognition and signal detection. Only when the class-conditional-probability density is known can the theory be used. A discretization method of stochastic variable (feature) space of the class-conditional-probability-density, and an estimation method for the class-conditional-probability-distribution are proposed. A Bayesian classification algorithm based on the methods is given. Finally, the methods are illustrated by applying them to radar target recognition

Keywords

Bayes methods; decision theory; pattern classification; probability; radar signal processing; radar target recognition; stochastic processes; Bayesian classifier; class-conditional-probability density; class-conditional-probability-distribution; discretization method; discretized continuous feature space; estimation method; pattern recognition; radar target recognition; signal detection; stochastic variable space; Bayesian methods; Function approximation; Multi-layer neural network; Neural networks; Pattern recognition; Probability density function; Space technology; Statistical distributions; Statistics; Stochastic processes;

fLanguage

English

Publisher

ieee

Conference_Titel

Signal Processing Proceedings, 1998. ICSP '98. 1998 Fourth International Conference on

Conference_Location

Beijing

Print_ISBN

0-7803-4325-5

Type

conf

DOI

10.1109/ICOSP.1998.770839

Filename

770839