Feature Weighted SVMs Using Receiver Operating Characteristics.

Support Vector Machines (SVMs) are a leading tool in classification and pattern recognition and the kernel function is one of its most important components. This function is used to map the input space into a high dimensional feature space. However, it can perform rather poorly when there are too many dimensions (e.g. for gene expression data) or when there is a lot of noise. In this paper, we investigate the suitability of using a new feature weighting scheme for SVM kernel functions, based on receiver operating characteristics (ROC). This strategy is clean, simple and surprisingly effective. We experimentally demonstrate that it can significantly and substantially boost classification performance, across a range of datasets. Key words: Receiver Operating Characteristics, Distance Function, Support Vector Machine, Classification.

• Type: Conference
• Year: 2009
• Where: SDM
• Authors: Shaoyi Zhang, M. Maruf Hossain, Md. Rafiul Hassan, James Bailey, Kotagiri Ramamohanarao
• Keywords: Computer Science, Feature Weighting Scheme, Receiver Operating Characteristics, SDM 2009, SVM Kernel Functions