Selecting subspace dimensions for kernel-based nonlinear subspace classifiers using intelligent search methods

Authors:
Sang-Woon Kim;B. John Oommen
Affiliations:
Senior Member, IEEE, Dept of Computer Science and Engineering, Myongji University, Yongin, Korea;Fellow of the IEEE, School of Computer Science, Carleton University, Ottawa, ON, Canada
Venue:
AI'04 Proceedings of the 17th Australian joint conference on Advances in Artificial Intelligence
Year:
2004

Citing 6
Cited 0

Nonlinear component analysis as a kernel eigenvalue problem

Neural Computation
Artificial Intelligence

Artificial Intelligence
Subspace Dimension Selection and Averaged Learning Subspace Method in Handwritten Digit Classification

ICANN 96 Proceedings of the 1996 International Conference on Artificial Neural Networks
On Utilizing Search Methods to Select Subspace Dimensions for Kernel-Based Nonlinear Subspace Classifiers

IEEE Transactions on Pattern Analysis and Machine Intelligence
Multi-category classification by kernel based nonlinear subspace method

ICASSP '99 Proceedings of the Acoustics, Speech, and Signal Processing, 1999. on 1999 IEEE International Conference - Volume 02
A formal analysis of why heuristic functions work

Artificial Intelligence

Quantified Score

Hi-index	0.00

Visualization

Abstract

In Kernel based Nonlinear Subspace (KNS) methods, the subspace dimensions have a strong influence on the performance ofthe subspace classifier In this paper, we propose a new method of systematically and efficiently selecting optimal, or near-optimal subspace dimensions for KNS classifiers using a search strategy and a heuristic function termed as the Overlapping criterion The task of selecting optimal subspace dimensions is equivalent to finding the best ones from a given problem-domain solution space We thus employ the Overlapping criterion of the subspaces as a heuristic function, by which the search space can be pruned to find the best solution to reduce the computation significantly Our experimental results demonstrate that the proposed mechanism selects the dimensions efficiently without sacrificing the classification accuracy The results especially demonstrate that the computational advantage for large data sets is significant.