Ensemble FARIMA Prediction with Stable Infinite Variance Innovations for Supermarket Energy Consumption

This paper concerns a fractional modeling and prediction method directly oriented toward an industrial time series with obvious non-Gaussian features. The hidden long-range dependence the multifractal property are extracted to determine order. A autoregressive integrated moving average model (FARIMA) is then proposed considering innovations stable infinite variance. existence convergence of solutions discussed in depth. Ensemble learning (ARMA) used further improve upon accuracy generalization. predict energy consumption real cooling system, superior results obtained.