Accession number:20084911757029
Title:Choquet integral logistic regression algorithm based on L- measure and γ-support
Authors:Liu, Hsiang-Chuan (1); Jheng, Yu-Du (2); Chen, Guey-Shya (2); Jeng, Bai-Cheng (2)
Author affiliation:(1) Department of Bioinformatics, Asia University Taiwan; (2) Graduate Institute of Educational Measurement and Statistics, Taichung University, Taiwan
Corresponding author:Liu, H.-C.
Source title:Proceedings of the 2008 International Conference on Wavelet Analysis and Pattern Recognition, ICWAPR
Abbreviated source title:Proc. Int. Conf. Wavelet Analysis and Pattern Recognition, ICWAPR
Volume:2
Monograph title:Proceedings of the 2008 International Conference on Wavelet Analysis and Pattern Recognition, ICWAPR
Issue date:2008
Publication year:2008 Pages:771-776
Article number:4635881 Language:English
ISBN-13:9781424422395
Document type:Conference article (CA)
Conference name:2008 International Conference on Wavelet Analysis and Pattern Recognition, ICWAPR
Conference date:August 30, 2008 - August 31, 2008 Conference location:Hong Kong, China
Conference code:74125
Publisher:Inst. of Elec. and Elec. Eng. Computer Society, 445 Hoes Lane - P.O.Box 1331, Piscataway, NJ 08855-1331, United States Abstract:Logistic regression algorithm and SVM algorithm are two well-known classification algorithms but when the multi-collinearity between independent variables occurs in above two algorithms, their classifying performance will always be not good. An improved classification algorithm combining the Choquet integral with respect to the λ-measure based on γ-support is proposed
by our previous work. In this paper, we replaced the more sensitive fuzzy measure, L-measure with the λ-measure in above improved classification algorithm, and we obtained a further improved algorithm, called Choquet integral logistic regression algorithm based on L-measure and γ-support. For
evaluating the performances of the SVM, logistic regression and the Choquet integral logistic regression algorithm with γ-
support based on P-measure, λ-measure and L-measure, respectively, a real data experiment by using Leave-one-out Cross- Validation accuracy is conducted. Experimental result shows that our new algorithm has the best performance. ©2008 IEEE.
Number of references:12 Main heading:Algorithms
Controlled terms:Feature extraction - Integral equations - Ketones - Logistics - Pattern recognition - Regression analysis - Support vector machines - Wavelet analysis - Wavelet transforms
Uncontrolled terms:Choquet integral - Choquet integrals -
Classification algorithms - Collinearity - Cross validations - Fuzzy measure - Fuzzy measures - Improved algorithms - Independent variables - L-measure - Leave one outs - Logistic regression
algorithms - Logistic regressions - New algorithms - Real datums - SVM algorithms
Classification code:922.2 Mathematical Statistics - 751.1 Acoustic Waves - 804.1 Organic Compounds - 912 Industrial Engineering and Management - 921 Mathematics - 921.2 Calculus - 921.3
Mathematical Transformations - 741.1 Light/Optics - 723.2 Data Processing and Image Processing - 723 Computer Software, Data Handling and Applications - 716 Telecommunication; Radar, Radio and Television - 703.2.1 Electric Filter Analysis - 404.1 Military Engineering - 723.5 Computer Applications
DOI:10.1109/ICWAPR.2008.4635881 Database:Compendex
Compilation and indexing terms, Copyright 2009 Elsevier Inc.
