The Bayesian method is firstly applied for the selection of the best subset for the double-threshold moving average (DTMA) model. The Markov chain Monte Carlo (MCMC) techniques and the stochastic search variable selection (SSVS) method are used to identify the best subset model from a very large number of possible models. Simulation experiments show that the proposed method is feasible and efficient, despite the complexity being increased by the large number of subsets, and the uncertainty of the threshold and delay variables. Our method is illustrated by real data analysis on the Yen-Dollar exchange rate.
|Number of pages||11|
|Journal||Statistics and Its Interface|
|Publication status||Accepted/In press - 12 Apr 2021|
- Bayesian estimation
- Monte Carlo Markov chains
- moving average models
- DTMA model