Title: Binary classification of functional data via continuously additive modeling
Authors: Kai Shen - North Carolina State University (United States)
Hans-Georg Mueller - University of California Davis (United States)
Yichao Wu - The University of Illinois at Chicago (United States) [presenting]
Fang Yao - Peking University, University of Toronto (China)
Abstract: The continuously additive model was recently proposed as a new nonlinear functional regression technique. It lends great flexibility to the study of functional data. To explore the use of continuously additive modeling for functional classification, we propose to couple it with the support vector machine, a large margin classifier, aiming at the classification of functional data. The support vector machine is a popular binary classification method that has enjoyed great success but has not become popular yet for the important task of classifying functional data. The support vector machine has been shown to be sensitive to outliers since it is based on an unbounded hinge loss. Therefore we couple the continuously additive modeling technique with the robust support vector machine using the truncated hinge loss as well. We illustrate the performance of our methods with simulation examples and two data sets that involve the classification of spectral data. The proposed approach is compared with classification based on the functional linear model.