Crude oil price forecasting using XGBoost

One of the most important role of economic variables in today's world countries are the price and the change of the price of crude oil. Changes in the price of crude oil have a very critical role in terms of treasury and budget, both in company and state planning. For example, one may choose one of the energy or natural gas indexed energy production plans based on the trend of the crude oil price, for planning to meet the need for electricity next year. Accurate forecasting of the crude oil price and realization of the forecasts based on this forecast will provide savings or gains in government and corporate economies, which can reach billions of dollars. There is a great need for this estimation in countries where crude oil production is low and heavily dependent on crude oil import. In this paper, the parameters which are the factors affecting the crude oil prices will be interpreted using XGBoost, a gradient boosting model, from machine learning libraries and estimation will be made.