On estimation of the effect lag of predictors on FDA models

Dr Haiyan Liu, University of Leeds. Part of the Statistics seminar series

We propose a functional linear model to predict a response using multiple functional and longitudinal predictors and to estimate the effect lags of predictors. The coefficient functions are written as the expansion of basis system (e.g. functional principal components, splines), and the coefficients of the fixed basis functions are estimated via optimizing a penalization criterion, then time lags are determined by simultaneously searching on a prior grid mesh based on minimization of prediction error criterion. Moreover, the mathematical properties of the estimated parameters and predicted responses are studied and the performance of the method is evaluated by extensive simulations. The analysis of Amazonian rainforest data not only illustrates the power of our model but also provides sound scientific interpretation for climate scientists and ecologists.