A discriminative training algorithm for predictive neural network models

Author

Na, KyungMin ; Rheem, JaeYeol ; Ann, Souguil

Author_Institution

Dept. of Electron. Eng., Seoul Nat. Univ., South Korea

Volume

6

fYear

1994

fDate

30 May-2 Jun 1994

Firstpage

431

Abstract

Predictive neural network models are powerful speech recognition models based on a nonlinear pattern prediction. Those models, however, suffer from poor discrimination between acoustically similar speech signals. In this paper, we propose a new discriminative training algorithm for predictive neural network models based on the generalized probabilistic descent (GPD) algorithm and the minimum classification error formulation. The proposed algorithm allows direct minimization of a recognition error rate. Evaluation of our training algorithm on Korean digits shows its effectiveness by 30% reduction of recognition error

Keywords

learning (artificial intelligence); neural nets; pattern classification; prediction theory; probability; speech recognition; Korean digits; acoustically similar speech signals; discriminative training algorithm; generalized probabilistic descent; minimum classification error formulation; nonlinear pattern prediction; predictive neural network models; recognition error rate; speech recognition models; Acoustical engineering; Artificial neural networks; Backpropagation algorithms; Dynamic programming; Neural networks; Power engineering and energy; Prediction algorithms; Predictive models; Speech recognition; Training data;

fLanguage

English

Publisher

ieee

Conference_Titel

Circuits and Systems, 1994. ISCAS '94., 1994 IEEE International Symposium on

Conference_Location

London

Print_ISBN

0-7803-1915-X

Type

conf

DOI

10.1109/ISCAS.1994.409618

Filename

409618