Stock Price Prediction on LQ45's Banking Sector Using Long-Short-Term Memory and Convolutional Neural Network

The stock market plays a crucial role in helping countries determine prices and ensuring liquidity. This study aims to develop a more effective method for predicting stock prices, focusing on the banking sector of the LQ45 (BBCA, BBRI, BBNI, BMRI, BBTN, and BRIS) due to its significant influence on the Indonesian market, with a market capitalization of 26.39%. By using Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) methods, this research compares their effectiveness in predicting stock prices using data from March 11, 2019, to March 11, 2024. The size of the data window is crucial in forecasting, and for this study, window sizes of 20 and 100 are used for analysis. The results indicate that LSTM shows superior performance. These findings highlight the potential of LSTM in capturing short-term trends for more accurate predictions in the banking sector, assisting investors and analysts in making well-informed investment decisions.