Title :
Efficient classification scheme based on hybrid global and local properties of feature
Author :
Lee, Heesung ; Hong, Sungjun ; An, Sungje ; Kim, Euntai
Author_Institution :
Sch. of Electr. & Electron. Eng., Yonsei Univ., Seoul
Abstract :
This paper proposes a new pattern classification scheme, combining global and local features. The proposed method uses principal component analysis (PCA) for global property and locality preserving projections (LPP) for local property of the pattern. PCA is known for preserving the most descriptive ones after projection while LPP is known for preserving the neighborhood structure of the data set. Our combing method integrates global and local descriptive information and finds a richer set of alternatives beyond PCA and LPP in a 2-D parametric space. In order to find the hybrid features adaptively and find optimal parameters, we employ the genetic algorithm (GA). Experiments are performed with UCI machine learning repository to show the performance of the proposed algorithm.
Keywords :
learning (artificial intelligence); pattern classification; principal component analysis; descriptive information; genetic algorithm; locality preserving projections; machine learning; pattern classification scheme; principal component analysis; Automatic control; Automation; Control systems; Genetic algorithms; Genetic mutations; Machine learning; Machine learning algorithms; Pattern classification; Principal component analysis; Robustness; Classification; GA; LPP; PCA; UCI machine learning repository;
Conference_Titel :
Control, Automation and Systems, 2008. ICCAS 2008. International Conference on
Conference_Location :
Seoul
Print_ISBN :
978-89-950038-9-3
Electronic_ISBN :
978-89-93215-01-4
DOI :
10.1109/ICCAS.2008.4694447