2014 : Forecasting Electricity Consumption in East Java-Bali Region Using Double Seasonal Exponential Smoothing and Double Seasonal ARIMA Model\n
Abstract
This research proposes a modeling approach for electricity consumption in East-Java \nBali Region. The first year research focused on applying Double Seasonal Holt Winter \nmethod and Double Seasonal ARIMA model. Both models are capable to capture more \nthan one seasonality order. Our preliminary investigation shows that double seasonality \nappears in the dataset i.e. daily and weekly seasonal. Considering the fact that the \nelectricity consumption growth with a certain degree of trend, Holt Winter approach \nwill be a good method to model this phenomena. Another advantage of Holt Winter \nmodel is its capability to model the seasonal pattern regardless of the degree of \nnonstationarity. It thus solves the problem of misspecification on the nonstationarity \ntreatment e.g. full differencing or fractional differenceing or long memory process. The \nresults of this approa