The study considered providing the best model among competing models for forecasting electricity load demand. Data from Helios Towers was used for this purpose. The study applied the Autoregressive Integrated Moving Average (ARIMA) model and residuals from the model tested for heteroscedasticity. The residuals were found to be heteroscedastic, thus, the Autoregressive Conditional Heteroscedastic (ARCH) and the Generalized Autoregressive Conditional Heteroscedastic (GARCH) models were applied to the data set. Competing models namely, ARCH (1), ARCH (2), ARCH (3) and GARCH (1, 1) models were fitted to the demand data under study. Based on the Akaike Information Criterion (AIC) and Bayesian Information Criterion (BIC) approach, the best heteroscedastic model was the ARCH (2). The study thus used the ARIMA (3, 0, 4)-ARCH (2) model for a two-point forecast of electricity load demand.
|Name||2021 International Conference on Cyber Security and Internet of Things (ICSIoT)|
|Conference||2021 International Conference on Cyber Security and Internet of Things|
|Period||15/12/21 → 17/12/21|