Error-Correcting Output Codes: A General Method for Improving Multiclass Inductive Learning Programs

Error-Correcting Output Codes: A General Method for Improving Multiclass Inductive Learning Programs (1991)

Cached

Download Links

[ftp.cis.ohio-state.edu]
[ftp.cs.orst.edu]

Other Repositories/Bibliography

DBLP

by Thomas G. Dietterich , Ghulum Bakiri

Venue:	IN PROCEEDINGS OF AAAI-91
Citations:	81 - 7 self

Versions

Version History

Metadata Version 3

User correction supplied by SystemCorrections

Datum	Value	Source
TITLE	Error-Correcting Output Codes: A General Method for Improving Multiclass Inductive Learning Programs	user correction - Legacy Corrections
AUTHOR NAME	Thomas Dietterich Ghulum	user correction - Legacy Corrections
AUTHOR NAME	Thomas G. Dietterich	user correction - Legacy Corrections
AUTHOR AFFIL	Department of Computer Science; Oregon State University	user correction - Legacy Corrections
AUTHOR ADDR	Corvallis, OR 97331-3202	user correction - Legacy Corrections
AUTHOR NAME	Ghulum Bakiri	user correction - Legacy Corrections
AUTHOR AFFIL	Department of Computer Science; Oregon State University	user correction - Legacy Corrections
AUTHOR ADDR	Corvallis, OR 97331-3202	user correction - Legacy Corrections
ABSTRACT	Multiclass learning problems involve finding a definition for an unknown function f(x) whose range is a discrete set containing k ? 2 values (i.e., k "classes"). The definition is acquired by studying large collections of training examples of the form hx i ; f(x i )i. Existing approaches to this problem include (a) direct application of multiclass algorithms such as the decision-tree algorithms ID3 and CART, (b) application of binary concept learning algorithms to learn individual binary functions for each of the k classes, and (c) application of binary concept learning algorithms with distributed output codes such as those employed by Sejnowski and Rosenberg in the NETtalk system. This paper compares these three approaches to a new technique in which BCH error-correcting codes are employed as a distributed output representation. We show that these output representations improve the performance of ID3 on the NETtalk task and of backpropagation on an isolated-letter speech-recognition t...	user correction - Legacy Corrections
YEAR	1991	INFERENCE
VENUE	In Proceedings of AAAI-91	INFERENCE
VENUE TYPE	CONFERENCE	INFERENCE
PAGES	572--577	INFERENCE
CITATIONS	19 found	ParsCit 1.0