Two-stage binary classifier with fuzzy-valued loss function

Authors:
Robert Burduk;Marek Kurzyński
Affiliations:
Department of Systems and Computer Networks, Wroclaw University of Technology, Wybrzeze Wyspianskiego 27, 50-370, Wroclaw, Poland;Department of Systems and Computer Networks, Wroclaw University of Technology, Wybrzeze Wyspianskiego 27, 50-370, Wroclaw, Poland
Venue:
Pattern Analysis & Applications
Year:
2006

Citing 0
Cited 5

Probability Error in Bayes Optimal Classifier with Intuitionistic Fuzzy Observations

ICIAR '09 Proceedings of the 6th International Conference on Image Analysis and Recognition
Randomness and fuzziness in bayes multistage classifier

HAIS'10 Proceedings of the 5th international conference on Hybrid Artificial Intelligence Systems - Volume Part I
Comparison of cost for zero-one and stage-dependent fuzzy loss function

ACIIDS'12 Proceedings of the 4th Asian conference on Intelligent Information and Database Systems - Volume Part I
Decomposition of classification task with selection of classifiers on the medical diagnosis example

HAIS'12 Proceedings of the 7th international conference on Hybrid Artificial Intelligent Systems - Volume Part II
Combining diverse one-class classifiers

HAIS'12 Proceedings of the 7th international conference on Hybrid Artificial Intelligent Systems - Volume Part II

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper we present the decision rules of a two-stage binary Bayesian classifier. The loss function in our case is fuzzy-valued and is dependent on the stage of the decision tree or on the node of the decision tree. The decision rules minimize the mean risk, i.e., the mean value of the fuzzy loss function. The model is first based on the notion of fuzzy random variable and secondly on the subjective ranking of fuzzy number defined by Campos and González. In this paper also, influence of choice of parameter λ in selected comparison fuzzy number method on classification results are presented. Finally, an example illustrating the study developed in the paper is considered.