Structured Multicategory Support Vector Machine with ANOVA Decomposition

• Activity Number: 111
• Type: Contributed
• Date/Time: Monday, August 9, 2004 : 10:30 AM to 12:20 PM
• Sponsor: Section on Statistical Graphics
• Author(s): Yoonkyung Lee*+ and Ja-Yong Koo and Yuwon Kim and Sangjun Lee
• Companies: Ohio State University and Inha University and Seoul National University and Seoul National University
• Keywords: Multicategory Support Vector Machine ; functional ANOVA decomposition ; feature selection ; classification ; 1-norm penalty

Abstract

The Support Vector Machine (SVM) has been a popular choice of classification method for many applications in machine learning. While it often outperforms other methods in terms of classification accuracy, the implicit nature of its solution renders the method less attractive in providing insights into the relationship between covariates and classes. Using structured kernels can remedy the drawback. Borrowing flexible model building idea of functional ANOVA decomposition, Multicategory Support Vector Machines with ANOVA kernels are considered in this paper. An additional penalty is imposed on the sum of weights of functional subspaces, which encourages a sparse representation of the solution. Incorporation of the additional penalty enhances the interpretability of a resulting classifier with often improved accuracy. The proposed method is demonstrated through simulation studies and applications to real data.