Research on Stock and Futures Price Forecasting Based on Machine Learning

Abstract

Impact on their decision-making process. Applying machine learning techniques to futures price forecasting can assist investors in making more rational investment decisions. Based on this, this paper selects the trading data of the "CSI 300 Futures" from 9:31:00 on January 3, 2017, to 15:00:00 on December 31, 2021, as a sample, and uses the Transformer model, Prophet model, and XGBoost model to predict and analyze its stock price trends. The final evaluation results show that the XGBoost model performs best in terms of prediction accuracy, while the Transformer model shows greater potential.