Functional data analysis with multi layer perceptrons

In this paper, we propose a way to apply multilayer perceptron (MLP) to functional data analysis. We introduce a computation model for functional input data and we show that this model is a well behaving extension of MLP: we show that the proposed model has the universal approximation property. Moreover, parameter estimation for this model is consistent. As a conclusion, we demonstrate functional MLP possibilities on simulated data and show they perform better than numerical MLP for a given number of parameters