Predicting the S&P 500 stock market with machine learning models

Predicting stock prices through machine learning models is becoming increasingly important in today's world. This essay aims to present a basic analysis of predicting the S&P 500 stock market using machine learning models, that are linear regression and XGBoost. The research utilizes historical data from the S&P 500 stock market in the past few years, dividing it into training, testing, and verification sets with a distribution of 70%, 15% and 15%, respectively. The linear regression model uses MSE (3.05110^17) and R^2(0.316) in testing group to analyze the accuracy of prediction, while the XGBoost model uses MSE (14816.886) and RMSE (121.725) in testing group to analyze the accuracy. The findings indicate that the accuracy of the XGBoost model surpasses that of the linear regression model. Drawing from the outcomes, one can infer that the XGBoost model is better suited for stock prediction compared to the linear regression model.