An Algorithm for Optimal Nonlinear Structure Preserving Feature Extraction.
RICE UNIV HOUSTON TEX DEPT OF ELECTRICAL ENGINEERING
Pagination or Media Count:
This paper presents a new approach to nonlinear structure preserving feature extraction. This method is based on certain graph theoretical considerations such as the minimal spanning tree, edge inconsistency, and diameter edges and topological considerations such as interpoint distance and measures. After introduction of the subject matter and appropriate background material, the algorithm is formulated. Numerical results from the application of this algorithm to various test data sets are presented.
- Statistics and Probability