Year: 2025
Type: article
Abstract: This paper discusses the shortcomings of traditional econometric methods in analyzing the nonlinear characteristics of the oil market, and proposes a new strategy based on AI and machine learning. We construct an analytical framework covering eight dimensions of finance, macroeconomics, inventory, policy uncertainty, technical indicators, supply and demand, and geopolitics, involving a total of 52 predictors. Through ADF test and differential processing, we ensure the stability of the data. Using the two-stage model, we screened out the key factors in 52 indicators through GCT, and further screened out 9 main indicators by RFE algorithm. Through KFold cross validation, we verify the robustness of this model and prove that it has a high degree of accuracy in predicting future oil prices and market fluctuations.
DOI: 10.54097/mcf8jf71
Authors: Guangjie Liu, Junjie Zhao, Jinping Zhang
Institution: Shenyang University of Chemical Technology
Language: English
Subfield: Economics and Econometrics
Domain: Social Sciences
Open Access status: diamond
OpenAlex ID: W4409621656