A New Approach for Forecasting Crude Oil Prices Based on Stochastic and Deterministic Influences of LMD Using ARIMA and LSTM Models

Crude oil is one of the non-renewable power sources and lifeblood contemporary industry. Every significant change in price crude (CO) will have an effect on how global economy, including COVID-19, develops. This study developed a novel hybrid prediction technique that depends local mean decomposition, Autoregressive Integrated Moving Average (ARIMA), Long Short-term Memory (LSTM) models to increase accuracy. The original data decomposed by decomposition (LMD), components are reconstructed into stochastic deterministic (SD) average mutual information reduce computation cost enhance forecasting accuracy, predict each individual component ARIMA, integrate residuals with LSTM capture nonlinearity help find final result. new model LMD-SD-ARIMA-LSTM has reduced volatility solved issue overfitting problem neural networks. proposed validated using publicly accessible from West Texas Intermediate (WTI), forecast accuracy compared measures. value Mean Absolute Error (MAE) Percentage (MAPE) for LSTM, LMD-ARIMA, LMD-SD-ARIMA, LMD-ARIMA-LSTM, LMD-SD-ARIMA-LSTM, Naïve 1.00, 1.539, 5.289, 0.873, 0.359, 0.106, 4.014 2.165, 1.832, 9.165, 1.359, 1.139, 1.124 3.821 respectively. From these results, it concluded minimum values MAE MAPE which assured superiority One-step ahead forecasting. Moreover, performance also up five steps ahead. findings demonstrate suggested approach helpful tool predicting CO prices both short long term. Furthermore, current reduces labor costs combing stationary non-stationary Product Functions (PFs) improved Meanwhile, traditional econometric can strengthen behavior after reconstruction, method better medium long-term price. accurate predictions provide reasonable advice relevant departments make correct decisions.