TY - JOUR AU - S. Folorunso AU - A. Taiwo AU - O. Olabanjo PY - 2019/03/19 Y2 - 2024/03/28 TI - A PREDICTIVE MODEL FOR ESTIMATING PETROLEUM CONSUMPTION USING MACHINE LEARNING APPROACH JF - LAUTECH Journal of Engineering and Technology JA - laujet VL - 12 IS - 2 SE - Articles DO - UR - https://laujet.com/index.php/laujet/article/view/297 AB - This study is focused on predicting the consumption of Petroleum (Thousands of Barrels per year) in Nigeria. Autoregressive integrated moving average (ARIMA), Linear Regression (LR) and Random Forest Regression (RFR) models were fitted to predict the consumption of Petroleum. The prediction accuracy of these models was evaluated using Root Mean Square Error (RMSE), Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE) and Coefficient of determination (R^2 ) metrics. The Petroleum dataset spanned a period of 37 years (1980-2017) and it was spilted into train and test at the ratio of 70:30 respectively to reduce overfitting. The result obtained revealed that the two machine learning models: LR and RFR outperformed the ARIMA model with lower values of prediction accuracy in terms of MAE, MAPE, RMSE and . ER -