Stock Price Prediction of High-tech Industry Based on LSTM and GRU: Tesla, Ferrari, and Walmart

Abstract

Stock price prediction has long been a critical area of focus both in the academic realm and in the practical world of finance, largely due to its direct influence on investment decisions and market dynamics. This study will specifically exam the stock prices of Tesla, Ferrari, and Walmart using advanced machine learning models, LSTM and GRU. Utilizing a rich dataset that spans several years, the study employs meticulous data preprocessing and feature engineering techniques, including the incorporation of external variables such as news sentiment and trading volume, to enhance prediction accuracy. Based on the results, the LSTM model outperforms the GRU model in accurately predicting the stock prices of Tesla and Walmart, while the GRU model shows a slight edge in forecasting Ferrari's stock prices. Furthermore, one finds that the inclusion of external features like news sentiment and trading volume considerably augments the predictive power of both LSTM and GRU models. These findings not only contribute to the knowledge hierarchy by providing a comparative analysis of LSTM and GRU in the context of the high-tech industry but also offer practical insights for investors, market analysts, and policymakers. The study serves as a robust framework for stock price prediction in volatile sectors and underscores the potential of integrating machine learning algorithms with feature engineering to achieve superior predictive performance.