Implementasi Metode LSTM untuk Prediksi Harga Saham PT Indofood CBP Sukses Makmur TBK

Abstract

High stock price fluctuations make stock prices difficult to predict accurately. Therefore, a predictive analysis approach that utilizes historical data in addition to machine learning methods is needed to help estimate price movements more effectively. This study aims to determine the performance of the Long Short-Term Memory (LSTM) method in predicting the stock prices of PT Indofood CBP Sukses Makmur Tbk based on historical data. LSTM is a type of artificial neural network that is effective in processing time series data due to its ability to capture long term relationships between data. Historical data is used to train the LSTM model. The results show that the LSTM model is effective in predicting stock prices, with an average accuracy of 80.5%. Sukses Makmur Tbk based on historical data. LSTM is one type of artificial neural network that is effective in processing time series data due to its ability to capture long-term relationships between data points. The data used consists of ICBP stock closing prices from January 2019 to May 2025. The methods used include data cleaning, data normalization, data splitting, model design, prediction, denormalization, and evaluation using the Mean Absolute Percentage Error (MAPE) metric. The research results demonstrate that the LSTM model performs well in recognizing time series data patterns, as indicated by the lowest MAPE value of 1.43, at the combination of 100 epochs and a batch size of 32.