Bayesian Inference for Smooth Transition Autoregressive Model

• Main Authors: Yu-Yun Chiang, 江昱昀
• Other Authors: Yu-Jau Lin
• Format: Others
• Language: zh-TW
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/9a7duw

碩士 === 中原大學 === 應用數學研究所 === 106 === The Bayesian statistics have been successfully commonly applied in many fields. The BUGS project, including OpenBUGS and its Windows version WinBUGS, has been one of the popular Bayesian soft. In BUGS language, users specify a statistical model by simply stating the likelihood function and the prior distributions of the corresponding parameters, then OpenBUGS computationally approximates the estimates of the empirical distribution by the MCMC methods. In this thesis, we consider to analyze the simulated data from smooth transition autoregressive (STAR) model using a newly developed R NIMBLE package, which is also defined in BUGS language and is much faster than OpenBUGS. Finally, the mean squared error of the estimates are computed to demonstrate the effectiveness of the MCMC method.