Applying Supervised Learning to Commodity Price Forecasting

Crude Oil, as a vital component of the world economy, requires precise price forecasting in order to control investment risks and maintain economic planning. Six supervised machine learning models—Decision Tree Regression (DTR), K-Nearest Neighbors (KNN), Multiple Linear Regression (MLR), Random Forest Regression (RFR), Support Vector Regression (SVR), and Multi-Layer Perceptron (MLP)—are used in this study to provide a thorough analysis of crude oil price prediction. Using Yahoo Finance's ten years of historical crude oil price data (2013–2023), we assess the model's performance at daily, weekly, and monthly intervals. Key financial indicators and related technical characteristics like EMA and MACD are included in the collection. Four statistical measures are used to evaluate models: R2 Score, Mean Absolute Percentage Error (MAPE), Root Mean Squared Error (RMSE), and Mean Squared Error (MSE). The findings show that DTR performs comparatively better on course-granularity data (monthly), whereas MLR and MLP consistently beat other models on daily and weekly data.