A FORECASTING METHODS FOR CO2 EMISSION FROM USING OIL: BOX - JENKINS METHOD AND WINTER’S METHOD

Abstract

This research aims to determine an optimum model and compares two methods for predicting the quantity of carbon dioxide from oil use including the Box-Jenkin method and the Winters method. The suitable model of Box-Jenkin method is ARIMA (0,1,1)(0,1,1)₁₂, and in the case of the Winters method, the forecasting model for the time t+1 from time t is _t (I) = (_t+I _t)_t+I-m , I = 1,2,…,m and _t (I) = (_t+I _t)_t+I-km , l = (k-1)m+1, (k-1)m+2, …,km; k = 1, 2, 3,… with the estimated parameters _t = 0.509 (Y_t / _t-m) + 0.491(_t-1 + _t-1), _t = 0.001 (_t -_t-1) + 0.999 _t-1, _t = 0.129 (Y_t / _t) + 0.871 _t-m We evaluate the correctness of the prediction using Mean Absolute Percentage Error (MAPE) and Root Mean Square Error (RMSE). Results indicate that MAPE and RMSE for the 3-month prediction are lower than those of the 5-moth prediction. The comparison of the two methods found that the Winters method produces a lower approximation error than the Box-Jenkin method in the 3-month prediction. In contrast, the Box-Jenkin method produces a lower approximation error than the Winters method in the case of 5-month prediction.