Bayesian Identification Procedure for Triple Seasonal Autoregressive Models

Triple seasonal autoregressive (TSAR) models have been introduced to model time series date with three layers of seasonality; however, the Bayesian identification problem these has not tackled in literature. Therefore, this paper, we objective filling gap by presenting a procedure identify best order TSAR models. Assuming that errors are normally distributed along employing priors, i.e., normal-gamma, Jeffreys’ and g on parameters, derive marginal posterior distributions parameters. In particular, show posteriors multivariate t gamma for coefficients vector precision, respectively. Using distribution vector, present an based sequence t-test significance. We evaluate accuracy proposed conducting extensive simulation study, followed real application hourly electricity load datasets six European countries.