Kernel-based Distance Metric Learning in the Output Space
Abstract
In this paper we present two related, kernel-based Distance Metric Learning (DML) methods. Their respective models non-linearly map data from their original space to an output space, and subsequent distance measurements are performed in the output space via a Mahalanobis metric. The dimensionality of the output space can be directly controlled to facilitate the learning of a low-rank metric. Both methods allow for simultaneous inference of the associated metric and the mapping to the output space, which can be used to visualize the data, when the output space is 2- or 3-dimensional. Experimental results for a collection of classification tasks illustrate the advantages of the proposed methods over other traditional and kernel-based DML approaches.
Cite
@article{arxiv.1312.2578,
title = {Kernel-based Distance Metric Learning in the Output Space},
author = {Cong Li and Michael Georgiopoulos and Georgios C. Anagnostopoulos},
journal= {arXiv preprint arXiv:1312.2578},
year = {2014}
}
Comments
11 pages, 7 figures, appeared in the Proceedings of 2013 International Joint Conference on Neural Networks (IJCNN)