COMPSTAT 2016: Start Registration
View Submission - CRoNoS FDA 2016
A0151
Title: On the range of integration of a functional linear model Authors:  Peter Hall - Melbourne University (Australia)
Giles Hooker - Cornell University (United States) [presenting]
Abstract: A conventional linear model for functional data involves expressing a scalar response variable in terms of a weighted integral of an explanatory functional covariate in which the weighting function is a parameter to be estimated. However, in some problems the support of this weight is a proper and unknown subset of the function's domain and is a quantity of particular practical interest. Motivated by a real-data example involving particulate emissions, we develop methods for estimating the upper end of this support along with other parameters in the functional linear model. We introduce techniques for selecting tuning parameters; and we explore properties of our methodology using both simulation and the real-data example mentioned above. Additionally, we derive theoretical properties of the methodology, and discuss implications of the theory. Our theoretical arguments give particular emphasis to the problem of identifiability.