Predicting the Stock Market

Abstract

This study examines whether historical stock data and moving averages improve the accuracy of LSTM models in predicting prices for technology and pharmaceutical companies. Using data from six major firms, the models were trained to identify patterns and forecast short-term price movements. The results show that while LSTM models can capture general trends, especially for more stable pharmaceutical stocks, their predictions are not consistently accurate, particularly for more volatile technology stocks. Short-term forecasts showed moderate success, but long-term predictions were unreliable due to error buildup and market unpredictability. Overall, the study concludes that historical data and moving averages help with trend detection but are not enough for precise stock prediction, as many external factors like news and economic events significantly influence stock prices.