Section: New Results

Model-based clustering for multivariate functional data

Participants : Julien Jacques, Cristian Preda.

We developed in [19] an extension of the model-based clustering algorithm for univariate functional data proposed in [20] , [23] , [11] to the case of multivariate functional data. For this, multivariate functional principal components analysis is defined and a parametric mixture model is proposed and estimated by an EM-like algorithm. Results on simulated and real datasets have shown the efficiency of the proposed method.