Adaptation of Gaussian ARD Kernel for Multiclass Classification

摘要：

The problem of optimizing Gaussian Automatic Relevance (ARD) kernel in a multiclass setting is considered. Unlike the conventional Gaussian kernel with a single width parameter, the Gaussian ARD kernel adopts multiple widths corresponding to the input features. We first present a model selection criterion named kernel distance-based class separability (KDCS) to evaluate the goodness of a kernel in multiclass classification scenario, then propose a gradient-based optimization algorithm to tune the width parameters of Gaussian ARD kernel via maximizing the KDCS criterion. This method is essentially a feature weighting method since each learned parameter indicates the relative importance of the corresponding feature. The proposed method is demonstrated with some UCI machine learning benchmark examples.

关键词： model selection feature weighting support vector machines (SVMs) auto relevance determination (ARD) multiclass classification

作者: Tinghua Wang

作者单位: School of Mathematics and Computer Science Gannan Normal University Ganzhou 341000, PR China

会议类型: 国际会议

会议名称: 2011 Eighth International Conference on Fuzzy System and Knowledge Discovery(第八届模糊系统与知识发现国际会议 FSKD 2011)

会议地点: 上海

会议语种:英文

页码: 1023-1026

在线出版日期: 2011-07-26（万方平台首次上网日期，不代表论文的发表时间）