Efficient resolution and basis functions selection in wavelet regression

Abstract

In this paper we propose an efficient method to determine a primary resolution and wavelet basis functions in wavelet regression. Most wavelet shrinkage methods focus on thresholding the wavelet coefficients, given a primary resolution which is usually determined by the sample size. However, both a primary resolution and the basis functions are affected by the shape of an unknown function rather than the sample size. Unlike existing methods, our method takes the shape of the unknown function into account because a proper resolution can be much affected by the shape of it rather than the sample size. Our approach to determine a primary resolution and wavelet basis functions is developed under Bayesian framework using the posterior model probability. We demonstrate the advantage of our approach using a simulation study and a real data application.