An efficient kernel matrix evaluation measure

Canh Hao Nguyen

doi:10.1016/j.patcog.2008.04.005

An efficient kernel matrix evaluation measure

2008-11-01

48

PRE

AI

DOI：10.1016/j.patcog.2008.04.005

原文链接

原文求助

摘要

En 中文

We study the problem of evaluating the goodness of a kernel matrix for a classification task. As kernel matrix evaluation is usually used in other expensive procedures like feature and model selections, the goodness measure must be Calculated efficiently. Most previous approaches are not efficient except for kernel target alignment (KTA) that can be calculated in O(n(2)) time complexity. Although KTA is widely used, we show that it has some serious drawbacks. We propose an efficient surrogate measure to evaluate the goodness of a kernel matrix based on the data distributions of classes in the feature space. The measure not only overcomes the limitations of KTA but also possesses other properties like invariance, efficiency and an error bound guarantee. Comparative experiments show that the measure is a good indication of the goodness of a kernel matrix. (C) 2008 Elsevier Ltd. All rights reserved.

Keyword:

classification

kernel methods

kernel matrix quality measure

kernel target alignment

class separability measure

An efficient kernel matrix evaluation measure

摘要

AI总结

期刊

机构

引用论文