Predicting the price of WTI crude oil futures using artificial intelligence model with chaos

This paper mainly investigates the chaotic features as well as prediction of WTI crude oil daily price time series from January 02, 2009 to November 25, 2019 with chaos theory and Artificial Intelligence technology, and analyzes the influence of noise on the price forecast. Firstly, we judge whether there is chaos in the time series of crude oil futures price by calculating the largest Lyapunov exponent, and applying Bootstrap technique to verify the robustness of the largest Lyapunov exponent. Secondly, we construct eight models with ANN technology and Chaos theory to fit the data and make short-term prediction. We find that the forecast accurateness can be improved dependent on the intrinsic formation mechanism (chaotic) model and removing noise can enhance the prediction accuracy of the model. Accurate prediction of WTI crude oil futures price is of high return for investors, low risk for risk management, and effective regulation of financial market for government departments. Finally, it is concluded that the EMD-LR-CHAOS model appeared to be the best prediction model among the eight models. As a first step, we calculate the greatest Lyapunov exponent and use the Bootstrap technique to test its robustness.