Forecasting oil production by adaptive neuro fuzzy inference system

Abstract

In this paper, the efficiency of neuro fuzzy network (ANFIS) is examined against auto regression (AR). Mean absolute percentage error (MAPE) is applied for this purpose. After applying different data preprocessing methods, the models are developed. A method for calculating ANFIS performance is also proposed. Due to various seasonal and monthly changes in oil production and difficulties in modeling it with conventional methods, we consider a case study in four countries for oil production estimation. Finally, analysis of variance (ANOVA) and Duncan Multiple Range Test (DMRT) is conducted for each country to evaluate the most efficient method.