GAUSSIAN PROCESS REGRESSION FOR FORECASTING GASOLINE PRICES IN JORDAN

The purpose of this paper is to forecast monthly gasoline prices in Jordan by applying Gaussian process regression on two types (octane-90 and octane-95) during the period January 2008 December 2019. Accurately predicting have several fiscal policy implications concerning fuel subsidies taxes. Also, they affect consumption production decisions. Moreover, are crucial for designing analyzing environmental policies. model was able treat a geometric Brownian motion with deterministic unknown drift function. performance prediction measured using Root Mean Square Error (RMSE) Average Percentage (MAPE). Where numerical results show that predictions were accurate. Keywords: Geometric motion; regression; Gasoline prices; Maximum likelihood method; Covariance function JEL Classifications: C11, C15, Q47, Q48 DOI: https://doi.org/10.32479/ijeep.11032