On preprocessing for mismatched classification of Gaussian signals

被引：2

作者：

Ephraim, Y ^{[1
]}

Roberts, WJJ

机构：

[1] George Mason Univ, Dept Elect & Comp Engn, Fairfax, VA 22030 USA

[2] George Mason Univ, Sch Informat Technol & Engn, Fairfax, VA 22030 USA

来源：

IEEE TRANSACTIONS ON INFORMATION THEORY | 2001年 / 47卷 / 03期

关键词：

hypothesis testing; mismatched classification; preprocessing;

D O I：

10.1109/18.915698

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

The optimal linear preprocessor for classifying two zero-mean Gaussian discrete-time signals which have been corrupted by additive zero-mean Gaussian noise is studied. Conditions for existence of the optimal linear preprocessor that achieves the performance of the Likelihood ratio test for the noisy signals are given and the preprocessor is explicitly derived.

引用

页码：1251 / 1256

页数：6

共 22 条

[1]

[Anonymous], 1971, DETECTION ESTIMATI 3

[2]

[Anonymous], TABLES INTEGRALS SER

[3]

BLAHUT RE, 1988, PRINCIPLES PRACTICES

[4] On second-order statistics and linear estimation of cepstral coefficients [J].

Ephraim, Y ;

Rahim, M .

IEEE TRANSACTIONS ON SPEECH AND AUDIO PROCESSING, 1999, 7 (02) :162-176

[5]

EPHRAIM Y, 2001, ENCY OPERATIONS RES

[6]

ESPOSITO R, 1968, INFORM CONTR, P116

[7]

GALLAGER RG, 1968, INFORMATION THEORY R

[8]

GRAY RM, 1998, 65041 STANF ELECT LA

[9]

JAFFER AG, 1972, INFORM CONTR, P46

[10] Detection of stochastic processes [J].

Kailath, T ;

Poor, HV .

IEEE TRANSACTIONS ON INFORMATION THEORY, 1998, 44 (06) :2230-2259

← 1 2 3 →