Undesirable effects of output normalization in multiple classifier systems

Authors:
Hakan Altinçay;Mübeccel Demirekler
Affiliations:
Department of Computer Engineering, Eastern Mediterranean University, Gazi Magusa, KKTC, Mersin 10, Turkey;Department of Computer Engineering, Eastern Mediterranean University, Gazi Magusa, KKTC, Mersin 10, Turkey and Department of Electrical and Electronics Engineering, Middle East Technical Universit ...
Venue:
Pattern Recognition Letters
Year:
2003

Citing 13
Cited 2

Decision Combination in Multiple Classifier Systems

IEEE Transactions on Pattern Analysis and Machine Intelligence
Democracy in neural nets: voting schemes for classification

Neural Networks
Combination of Multiple Classifiers Using Local Accuracy Estimates

IEEE Transactions on Pattern Analysis and Machine Intelligence
On Combining Classifiers

IEEE Transactions on Pattern Analysis and Machine Intelligence
Lipreading: a classifier combination approach

Pattern Recognition Letters - special issue on pattern recognition in practice V
Adaptive confidence transform based classifier combination for Chinese character recognition

Pattern Recognition Letters
Combining multiple matchers for a high security fingerprint verification system

Pattern Recognition Letters - Special issue on pattern recognition in practice VI
An information theoretic framework for weight estimation in the combination of probabilistic classifiers for speaker identification

Speech Communication
Person Identification Using Multiple Cues

IEEE Transactions on Pattern Analysis and Machine Intelligence
Classifier Conditional Posterior Probabilities

SSPR '98/SPR '98 Proceedings of the Joint IAPR International Workshops on Advances in Pattern Recognition
Methods for Dynamic Classifier Selection

ICIAP '99 Proceedings of the 10th International Conference on Image Analysis and Processing
Pattern Classification (2nd Edition)

Pattern Classification (2nd Edition)
Information combination operators for data fusion: a comparative review with classification

IEEE Transactions on Systems, Man, and Cybernetics, Part A: Systems and Humans

Post-processing of Classifier Outputs in Multiple Classifier Systems

MCS '02 Proceedings of the Third International Workshop on Multiple Classifier Systems
Multimodal biometrics: issues in design and testing

Proceedings of the 5th international conference on Multimodal interfaces

Quantified Score

Hi-index	0.10

Visualization

Abstract

Incomparability of the classifier output scores is a major problem in the combination of different classification systems. In order to deal with this problem, the measurement level classifier outputs are generally normalized. However, empirical results have shown that output normalization may lead to some undesirable effects. This paper presents analyses for some most frequently used normalization methods and it is shown that the main reason for these undesirable effects of output normalization is the dimensionality reduction in the output space. An artificial classifier combination example and a real-data experiment are provided where these effects are further clarified.