NONPARAMETRIC BAYESIAN SUPERVISED CLASSIFICATION OF FUNCTIONAL DATA

A nonparametric approach combining generative models and func-tional data analysis is presented in this paper for classifying func-tional data which arise naturally in a wide variety of signal process-ing applications, such as brain computer interfacing, speech recog-nition, or image classification. Based on a new and improved family of Bayesian classifiers, we extend hierarchical Bayesian classifica-tion methodology from vector to functional settings. We provide theoretical and practical motivations to our approach which relies on Dirichlet process mixtures and Gaussian processes. The perfor-mance is evaluated on phoneme recognition task, and compared to that of Functional Support Vector Machines (FSVMs). Index Terms — Functional data analysis, supervised classifica