DOS ENFOQUES PARA EL ANÁLISIS DE COMPONENTES PRINCIPALES FUNCIONALES MÚLTIPLES: UN ESTUDIO COMPARATIVO

Abstract

Functional principal component analysis (FPCA) is a technique which has been gaining power in applications by means of getting acquainted with its scope and the possibilities of interpretation of complex data. A direct generalization of FPCA is the Multiple Functional Principal Component Analysis (MFPCA) which deals with multiple time series. The amplest known approach is due to Ramsay and Silverman (2002). In this paper the proposal by Berrendero et al. (2011) is presented and a comparative study is done and a linking strategy is proposed in search of a better quality in the interpretation of the results. This strategy is applied to a socioecological study related with the evolution of lobster fishing in the western part of Cuba.