Semiparametric shape-invariant models for periodic data

Abstract

This article presents a novel shape-invariant modeling approach to quasi-periodic data. We propose a dynamic semiparametric method that estimates the common cycle shape in a nonparametric way and the individual phase and amplitude variability in a parametric way. An efficient algorithm to compute the estimators is proposed. The behavior of the estimators is studied by simulation and by a real-data example.

Keywords:

• circadian rhythms
• nonparametric regression
• spline smoothing