Variational Bayes for Regime-Switching Log-Normal Models

Entropy

Hui Zhao¹ 

[1] University of Waterloo, 200 University Avenue West, Waterloo, ON N2L 3G1, Canada; E-Mail

关键词: information geometry;    variational Bayes;    regime-switching log-normal model;    model selection;    covariance estimation;   

DOI  :  10.3390/e16073832

来源: mdpi

【 摘 要 】

The power of projection using divergence functions is a major theme in information geometry. One version of this is the variational Bayes (VB) method. This paper looks at VB in the context of other projection-based methods in information geometry. It also describes how to apply VB to the regime-switching log-normal model and how it provides a computationally fast solution to quantify the uncertainty in the model specification. The results show that the method can recover exactly the model structure, gives the reasonable point estimates and is very computationally efficient. The potential problems of the method in quantifying the parameter uncertainty are discussed.

【 授权许可】

CC BY   
© 2014 by the authors; licensee MDPI, Basel, Switzerland