A1747
Title: F-type tests for functional linear models
Authors: Shu-Chin Lin - National Taiwan University (Taiwan) [presenting]
Abstract: Hypothesis testing for functional linear regression models with a scalar response is considered under the assumption that the coefficient function resides in a reproducing kernel Hilbert space. The goal is to test whether the coefficient function belongs to a closed subspace of the reproducing kernel Hilbert space, providing a unified platform to test global nullity, partial nullity, and parametric forms for the coefficient function. A simple and efficient testing procedure is developed based on a statistic that resembles the classical F-test. The minimax rate for the testing problems is studied, and the proposed testing procedure is shown to be asymptotically close to the optimal setting under mild conditions.
