Stock Price Prediction of Apple Inc. Based on LSTM Model: An Application of Artificial Intelligence in Individual Stock Analysis

This paper explores the potential of artificial intelligence in financial forecasting by applying a Long Short-Term Memory (LSTM) neural network to the task of predicting Apple Inc. (AAPL) stock prices. Using historical daily closing prices from August 2020 to August 2023, the model was trained and tested under a structured framework of data preprocessing, normalization, and sequential input construction. The forecasting performance was evaluated with widely used error measures, including Mean Squared Error (MSE), Mean Absolute Error (MAE), and Root Mean Squared Error (RMSE). Experimental results indicate that the LSTM model effectively captures sequential dependencies and nonlinear dynamics within the time series, generating predictions that closely align with observed prices. While the model demonstrates high accuracy in relatively stable market conditions, its reliance on univariate input limits adaptability to abrupt market fluctuations and external influences such as macroeconomic shifts. These findings suggest that LSTM-based models can serve as valuable tools for supporting individual stock analysis, but their effectiveness depends on stock-specific data availability and market characteristics. Future work may extend this framework by incorporating multi-source financial indicators and enhancing model interpretability to achieve broader practical relevance.