Simulating classifier outputs for evaluating parallel combination methods

Authors:
H. Zouari;L. Heutte;Y. Lecourtier;A. Alimi
Affiliations:
Laboratoire Perception, Systèmes, Information, Université de Rouen, France;Laboratoire Perception, Systèmes, Information, Université de Rouen, France;Laboratoire Perception, Systèmes, Information, Université de Rouen, France;Groupe de Recherche sur les Machines Intelligentes, Université de Sfax, Tunisie
Venue:
MCS'03 Proceedings of the 4th international conference on Multiple classifier systems
Year:
2003

Citing 5
Cited 2

Decision Combination in Multiple Classifier Systems

IEEE Transactions on Pattern Analysis and Machine Intelligence
On Combining Classifiers

IEEE Transactions on Pattern Analysis and Machine Intelligence
Generating classifier outputs of fixed accuracy and diversity

Pattern Recognition Letters
Evaluation of Combination Methods

ICDAR '99 Proceedings of the Fifth International Conference on Document Analysis and Recognition
An Overview and Comparison of Voting Methods for Pattern Recognition

IWFHR '02 Proceedings of the Eighth International Workshop on Frontiers in Handwriting Recognition (IWFHR'02)

Rapid and brief communication: Controlling the diversity in classifier ensembles through a measure of agreement

Pattern Recognition
An Iterative Method for Deciding SVM and Single Layer Neural Network Structures

Neural Processing Letters

Quantified Score

Hi-index	0.00

Visualization

Abstract

The use of artificial outputs generated by a classifier simulator has recently emerged as a new trend to provide an underlying evaluation of classifier combination methods. In this paper, we propose a new method for the artificial generation of classifier outputs based on additional parameters which provide sufficient diversity to simulate, for a problem of any number of classes and any type of output, any classifier performance. This is achieved through a two-step algorithm which first builds a confusion matrix according to desired behaviour and secondly generates, from this confusion matrix, outputs of any specified type. We provide the detailed algorithms and constraints to respect for the construction of the matrix and the generation of outputs. We illustrate on a small example the usefulness of the classifier simulator.