Forecast of the energy consumption of Turkiye commerce sector: m-estimation model application

Abstract

Net electricity consumption use continue to be significant issues. There are various forms of energy use and production. This work uses robust and form of Mestimation by using the grid-search algorithm. Thus, since we use form of Huber Mestimation, the prediction performance can be increased; because, the data is tried to be modelled by using the different values of tuning parameter determined by the gridsearch which can be used to carry out the optimization proving the M-estimates of the parameters of regression model. The data sets which are year and the net electricity consumption are modelled by regression model in order to predict and forecast how much electricity consumption will be necessary for the commercial purpose firm used the electricity at the highest amount. The statistical inference for the regression model and its estimators of parameters in the model is also provided. Further, the illustrative results used for the grid-search and the analytical expression of regression model are given. Due to the fact that polynomial regression showing an increment in the polynomial trend can model the dependent variable well, the net electrical consumption in commerce at Turkiye increases and the bandwidths for the forecasting in the year 2021 and 2022 are given to conduct a planning in energy sector.