Ervin Yohannes, Aldin Febriansyah, Nisa Dwi Septiyanti, Suparji, Agus Wiyono, Aries Dwi Indriyanti, Fitri Utaminingrum, Chih-Yang Lin, Kahlil Muchtar, Avirmed Enkhbat
Investors often face significant challenges in predicting fluctuating stock price movements, which can lead to uncertainty and suboptimal investment decisions. This study aims to evaluate the performance of the Long Short-Term Memory (LSTM) deep learning model in forecasting stock prices. The dataset utilized is derived from the Pakistan Stock Exchange (KSE 100), and the model's performance is assessed using evaluation metrics including Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and Mean Absolute Percentage Error (MAPE). The experimental results demonstrate that the LSTM model achieves strong predictive performance, with the lowest recorded MSE of 0.0004, RMSE of 0.0210, MAE of 0.0141, and MAPE of 0.0209, corresponding to an accuracy rate of 97.91%. These findings highlight the effectiveness of the LSTM model in stock price prediction and provide valuable insights for investors seeking to enhance decision-making through data-driven forecasting approaches. © 2025 IEEE.
State University of Surabaya, Surabaya, Indonesia; State University of Surabaya, Surabaya, Indonesia; State University of Surabaya, Surabaya, Indonesia; State University of Surabaya, Surabaya, Indonesia; State University of Surabaya, Surabaya, Indonesia; University of Brawijaya, Malang, Indonesia; National Central University, Taoyuan, Taiwan; Syiah Kuala University, Banda Aceh, Indonesia; National University of Mongolia, Mongolia
