How Well do Bayes Methods Work for On-Line Prediction of {&plusmn;1} values?

How Well do Bayes Methods Work for On-Line Prediction of {±1} values? (1992)

Cached

Download Links

[ftp.cse.ucsc.edu]

by D. Haussler , A. Barron

Venue:	In Proceedings of the Third NEC Symposium on Computation and Cognition. SIAM
Citations:	17 - 10 self

Versions

Version History

Metadata Version 3

User correction supplied by SystemCorrections

Datum	Value	Source
TITLE	How Well do Bayes Methods Work for On-Line Prediction of {±1} values?	user correction - Legacy Corrections
AUTHOR NAME	D. Haussler	SVM HeaderParse 0.1
AUTHOR AFFIL	Baskin Center for Computer Engineering and Information Sciences; University of California, Santa Cruz	SVM HeaderParse 0.2
AUTHOR ADDR	CA 95064	SVM HeaderParse 0.1
AUTHOR NAME	A. Barron	SVM HeaderParse 0.1
AUTHOR AFFIL	; Dept. of Statistics, U. Ill.; Computer and Information Sciences; University of California at Santa Cruz	SVM HeaderParse 0.2
AUTHOR ADDR	; 713 Wright Street; Champaign IL 61820; sees it [HLW90, Lit89, LW89, Vov90b, HKS91, OH91a, SST92, MF92]. Such systems adapt	SVM HeaderParse 0.1
ABSTRACT	We look at sequential classification and regression problems in which f\Sigma1g-labeled instances are given on-line, one at a time, and for each new instance, before seeing the label, the learning system must either predict the label, or estimate the probability that the label is +1. We look at the performance of Bayes method for this task, as measured by the total number of mistakes for the classification problem, and by the total log loss (or information gain) for the regression problem. Our results are given by comparing the performance of Bayes method to the performance of a hypothetical "omniscient scientist" who is able to use extra information about the labeling process that would not be available in the standard learning protocol. The results show that Bayes methods perform only slightly worse than the omniscient scientist in many cases. These results generalize previous results of Haussler, Kearns and Schapire, and Opper and Haussler. 1 Introduction Several recent papers in...	user correction - Legacy Corrections
YEAR	1992	INFERENCE
VENUE	In Proceedings of the Third NEC Symposium on Computation and Cognition. SIAM	INFERENCE
VENUE TYPE	CONFERENCE	INFERENCE
PAGES	74--100	INFERENCE
CITATIONS	36 found	ParsCit 1.0

How Well do Bayes Methods Work for On-Line Prediction of {&plusmn;1} values? (1992)

Cached

Download Links

Versions

Version History

Metadata Version 3

How Well do Bayes Methods Work for On-Line Prediction of {±1} values? (1992)