Application of machine learning model in stock prediction

Abstract

Economic stability and investor decision-making are significantly impacted by the stock market, which plays a crucial role in financial markets worldwide. This study focuses on applying machine learning models, particularly Random Forest, Extreme Gradient Boosting (XGBoost), and Long Short-Term Memory (LSTM), to predict Tesla stock prices. Various models' performance in stock prediction is evaluated using R-squared (R2), Mean Absolute Error (MAE), and Root Mean Square Error (RMSE). The results show that Random Forest and LSTM achieve the highest forecast accuracy, outperforming other models. These findings suggest that machine learning models are effective tools for stock prediction, providing valuable insights for investors. The superior performance of Random Forest and LSTM may be attributed to their ability to capture complex patterns in the data, especially in volatile markets. To further enhance the precision and reliability of stock predictions, future research should explore multi-source data integration and advanced model optimization techniques. Additionally, incorporating market sentiment and macroeconomic indicators could improve prediction robustness and offer deeper insights into market trends.