Gold Price Prediction using Long Short-Term Memory Algorithm

Abstract

Gold is one of the most popular investment instruments among the public because it has low risk, stable value, and is resistant to inflation.  In Indonesia, ANTAM gold attracts strong demand due to its authenticity and relative price stability; however, frequent supply limitations and external economic factors cause price fluctuations that complicate investment timing decisions. In this paper, we utilize a data-driven analytical approach to model and predict gold price movements by applying the Long Short-Term Memory (LSTM) algorithm, which is well-suited for capturing temporal dependencies in time series data. We adopt historical ANTAM gold price data to train and evaluate the model, allowing it to learn underlying price patterns and long-term trends effectively. The experimental results demonstrate that the proposed LSTM model achieves high predictive accuracy, as reflected by an RMSE of 0.021781 and a MAPE of 1.94% when trained with 100 epochs, indicating a very low average prediction error. These findings confirm that the LSTM-based approach is effective for forecasting precious metal gold prices and shows strong potential as a practical decision-support tool for short- and long-term investment planning.