T-Test model for context aware classifier

Authors:
Mi Young Nam;Battulga Bayarsaikhan;Suman Sedai;Phill Kyu Rhee
Affiliations:
Dept. of Computer Science & Engineering, Inha University, Incheon, Korea;Dept. of Computer Science & Engineering, Inha University, Incheon, Korea;Dept. of Computer Science & Engineering, Inha University, Incheon, Korea;Dept. of Computer Science & Engineering, Inha University, Incheon, Korea
Venue:
ICNC'06 Proceedings of the Second international conference on Advances in Natural Computation - Volume Part I
Year:
2006

Citing 5
Cited 0

A Theoretical Study on Six Classifier Fusion Strategies

IEEE Transactions on Pattern Analysis and Machine Intelligence
Reconfigurable Context-Sensitive Middleware for Pervasive Computing

IEEE Pervasive Computing
Development of Situation-Aware Application Software for Ubiquitous Computing Environment

COMPSAC '02 Proceedings of the 26th International Computer Software and Applications Conference on Prolonging Software Life: Development and Redevelopment
Designing classifier fusion systems by genetic algorithms

IEEE Transactions on Evolutionary Computation
Switching between selection and fusion in combining classifiers: anexperiment

IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper proposes a t-test decision model for context aware classifier combination scheme based on the cascade of classifier selection and fusion. In the proposed scheme, system working environment is learned and the environmental context is identified. Best selection is applied to the environment context where one classifier strongly dominates the other. In the remaining context, fusion of multiple classifiers is applied. The decision of best selection or fusion is made using t-test decision model. Fusion methods namely Cosine based identify and Euclidian identify. In the proposed scheme, we are modeling for t-test based combination system. A group of classifiers are assigned to each environmental context in prior. Then the decision of fusion of more than one classifiers or selecting best classifier is made using proposed t-test decision model.