<![CDATA[Gold price fluctuations pose significant challenges for investors in determining accurate investment strategies. The volatility is strongly influenced by inflation, exchange rates, and global economic dynamics, making reliable forecasting increasingly important. Although various statistical and machine learning models have been applied, many are limited in capturing complex temporal dependencies, especially in the context of Indonesia’s ANTAM gold prices. This study addresses that gap by applying the Long Short-Term Memory (LSTM) method, a deep learning approach designed to model sequential patterns in time series data. The novelty of this research lies in the application of LSTM specifically for ANTAM gold price forecasting in Indonesia, which has received limited attention in previous studies. Unlike conventional approaches, LSTM is capable of preserving long-term dependencies, thereby improving predictive accuracy for volatile commodities. Using historical daily data from November 2023 to March 2025, the model was trained to recognize price dynamics and evaluated with Root Mean Square Error (RMSE) and Mean Absolute Percentage Error (MAPE). The results demonstrate high predictive accuracy, with a MAPE of 1.39% and RMSE of 0.0137. These findings confirm the suitability of LSTM for gold price prediction and underline its potential contribution to both theoretical advancements in time series forecasting and practical decision-making in investment management. Thus, this study not only strengthens evidence of LSTM’s effectiveness but also offers valuable insights for investors and policymakers in managing risks associated with commodity price volatility.]]>
