A Comparison of Testing Methods in Scalar-on-Function Regression

• DOI: 10.48550/arxiv.1710.05729
• Publication Date: 2017-01-01
• Authors: Merve Yasemin Tek...; Marcela Alfaro‐Có...; Arnab Maity; Ana‐Maria Staicu
• Keywords: Methodology (stat.ME); FOS: Computer and information sciences; 62G10, 62G08, 62F03, 62J99; G.3; Statistics - Methodology

A scalar-response functional model describes the association between a scalar response and set of covariates. An important problem in data literature is to test nullity or linearity effect covariate context scalar-on-function regression. This article provides an overview existing methods for testing both null hypotheses that there no relationship linear response, comprehensive numerical comparison their performance. The are compared variety realistic scenarios: when observed at dense sparse grids measurements include noise not. Finally, illustrated on Tecator set.