DocumentCode :
3059402
Title :
Learning class probabilities from labeled data
Author :
Singer, Yoram ; Yair, Eyal
Author_Institution :
IBM-Sci. & Technol., Technion City, Haifa, Israel
fYear :
1992
fDate :
30 Aug-3 Sep 1992
Firstpage :
553
Lastpage :
556
Abstract :
A Bayesian classifier may supply an optimal estimate of the a posteriori class probabilities for classifying stochastic patterns, provided that the underlying statistical model of the problem is known. In the absence of such a priori knowledge, one valuable alternative is the Boltzmann perceptron classifier (BPC), a statistical neural based classifier, which was shown to have the capability of Bayesian like decisions. The original learning algorithm of the BPC requires a knowledge of the a posteriori probabilities for the given training set. However, these probabilities are seldom known in advance, and instead, labeled training data is given for which only the class membership associated with each training sample is known. The authors introduce a regulated learning scheme which estimates the class probabilities from such labeled data and constructs a classifier that generalizes well for new data
Keywords :
Boltzmann machines; learning (artificial intelligence); pattern recognition; probability; statistics; Boltzmann perceptron classifier; class membership; class probabilities; labeled training data; learning algorithm; statistical neural based classifier; statistical pattern recognition; Bayesian methods; Cities and towns; Computer architecture; Pattern recognition; Probability; Robustness; Stochastic processes; Training data;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Pattern Recognition, 1992. Vol.II. Conference B: Pattern Recognition Methodology and Systems, Proceedings., 11th IAPR International Conference on
Conference_Location :
The Hague
Print_ISBN :
0-8186-2915-0
Type :
conf
DOI :
10.1109/ICPR.1992.201839
Filename :
201839
Link To Document :
بازگشت