“…When handling time series with linear characteristics, its forecasting performance is generally not as good as the ARIMA model's (Maia, de Carvalho & Ludermir, 2008). Lee, Wang, Hsu and Lai (2005) used the non-informative diffuse-prior Bayesian vector autoregression model to forecast the productions of technology industries. However, it is not suitable for univariate time series.…”