Bayesian Inference of Triple Seasonal Autoregressive Models

In this paper we extend autoregressive models to fit time series that have three layers of seasonality, i.e. triple seasonal (TSAR) models, and introduce the Bayesian inference for these TSAR models. Assuming model errors are normally distributed employing priors, Jeffreys', g, normal-gamma on parameters, derive marginal posterior distributions parameters. particular, show be multivariate t gamma coefficients precision, respectively. We evaluate efficiency proposed using simulation study, then apply it real-world hourly electricity load datasets in six European countries.