COMPSTAT 2018: Start Registration
View Submission - COMPSTAT2018
Title: A multiple functional linear model Authors:  Dominique Katshunga - University of Cape Town (South Africa) [presenting]
Abstract: The increased need to base statistical models on good quality data has led to the approaches of functional data and functional data analysis (FDA). Functional data refer to data providing information about curves and surfaces varying over a continuum, with smoothness being the key feature. Functional data analysis has enjoyed a considerable amount of attention in the literature under its exploratory and confirmatory statistical analyses approaches. However, under the predictive approach, functional data analysis is still an emerging field. As a part of this emerging field, this discussion focusses on extending the functional linear models (FLM) for functional responses to accommodate more than one covariates. Emphasis is put on fitting the extended models using the backfitting technique. As an illustration, a simulation study is considered.