A0845
Title: Truncated estimation for functional linear model and its application to agricultural data
Authors: Hidetoshi Matsui - Shiga University (Japan) [presenting]
Abstract: Truncated estimation for the functional linear model is a useful technique for investigating the relationship between a functional predictor and a scalar response. We consider the problem of estimating a varying-coefficient functional linear model, where the predictor is a function of time and the scalar response depends on not only a functional predictor but also an exogenous variable. The aim is to estimate the model so that the functional predictor does not relate to the response after a certain point in time at any value of the exogenous variable. We apply the sparse regularization to shrink the corresponding domain of the coefficient function towards exactly zero. Simulation studies are conducted to investigate the effectiveness of the proposed method. We also apply the method to the analysis of agricultural data to identify when an environmental factor relates to the crop yield.
