Bayesian empirical likelihood inference and order shrinkage for a hysteretic autoregressive model
摘要
In this article, we consider Bayesian empirical likelihood (BEL) inference for a class of hysteretic autoregressive models. The primary focus of this study is to develop a BEL method that integrates Bayesian inference and the empirical likelihood method for estimating the hysteretic autoregressive (HAR) model. Additionally, the order determination problem of the HAR model is discussed under the prior assumption of spike-and-slab. We conduct simulation studies to illustrate the advantages of the proposed BEL method and apply it to analyzing the U.S. Industrial Production Index data set.