Variable Selection for Classification of Multivariate Functional Data

Abstract

New variable selection method is considered in the setting of classification with multivariate functional data (Ramsay and Silverman, Functional data analysis, 2005). The variable selection is a dimensionality reduction method which leads to replace the whole vector process, with a low-dimensional vector still giving a comparable classification error. The various classifiers appropriate for functional data are used. The proposed variable selection method is based on functional distance covariance (Szekely et al. Ann Appl Stat 3(4):1236–1265, 2009; Stat Probab Lett 82(12):2278–2282, 2012). and is a modification of the procedure given by Kong et al. (Stat Med 34:1708–1720, 2015). The proposed methodology is illustrated on real data example.