Linear methods for regression and classification with functional data
Gilbert Saporta, Damiana Costanzo, Cristian Preda · HAL (Le Centre pour la Communication Scientifique Directe) · 2007
Functional data occurs when we observe curves or paths from a stochastic process Xt .If for each curve or path we have a single response variable Y, we have a regression problem when Y is numerical, a classification problem when Y is categorical. We assume here that all trajectories are observed continuously on a time interval [0;T] and that the variables Y (when numerical) and Xt have zero mean.