A0882
Title: Estimation for multiple-threshold regression models
Authors: ChihHao Chang - National University of Kaohsiung (Taiwan) [presenting]
Abstract: Threshold linear regression models with multiple threshold points are considered. We provide algorithms to efficiently estimate threshold points, where the number of thresholds and the type of thresholds (kink or jump) can be detected simultaneously. We conduct several simulation studies to demonstrate the finite-sample performance of the proposed algorithm versus existing algorithms. The simulation results reveal that the our method has less computational burden, comparable estimation accuracy and better flexibility for threshold type detection.
