Non-linear Feature Extraction using Generic Radial Projections

A.J. Perez-Jimenez and J.C. Perez-Cortes (Spain)

Keywords

feature extraction, clustering, Mahalanobis distance, genetic algorithms.

Abstract

In this work, a new technique for non-linear feature extraction is presented. In this technique, new features are obtained as Generic Radial Projections of the original measurements. Radial projections are second order trans formations that capture the local structure of the data. A number of results of the proposed method are presented in this work.

Important Links:



Go Back