|
Widely applicable information criterion for estimatingthe order in a hidden Markov model |
|---|---|
| รหัสดีโอไอ | |
| Creator | 1. Safaa K. Kadhem 2. Sadeq A. Kadhim |
| Title | Widely applicable information criterion for estimatingthe order in a hidden Markov model |
| Publisher | Research and Development Office,Prince of Songkla University |
| Publication Year | 2021 |
| Journal Title | Songklanakarin Journal of Science and Technology (SJST) |
| Journal Vol. | 43 |
| Journal No. | 3 |
| Page no. | 824-833 |
| Keyword | hidden Markov chains models, Markov chain Monte Carlo, integrated posterior predictive density, model selection |
| URL Website | https://rdo.psu.ac.th/sjstweb/ |
| ISSN | 0125-3395 |
| Abstract | This paper considers the determination of the order of hidden Markov models. Recently, a proposed predictivemeasure, the so-called widely applicable information criterion (WAIC), was derived. This criterion is a convenient alternative tothe cross-validation approach due to its less computation processes and quick evaluation. We studied the properties of thiscriterion applied to hidden Markov models (HMMs) under the Bayesian principle. Such models include serial dependence andoverdispersion of observed data. We investigated this criterion via simulation studies and a real data application. It is shown thatthe introduced criterion performs better with less complicated models, while it tends to over fit some more complicated models. |
