<![CDATA[The Southern part of Sumatra is highly vulnerable to earthquakes due to its location in the subduction zone between the Indo-Australian plate and the Sunda plate. The Southern part of Sumatra’s vulnerability to earthquakes poses significant risks. This research aims at predicting the Earthquakes in the Southern part of Sumatra Using Deep Learning (Long Short-Term Memory) Models, a deep learning method designed to analyze sequential data. The model utilized 20 years of historical earthquake data from 2004 to 2024, with parameters including magnitude, epicenter location, depth, and event time. Data were preprocessed using Min-Max Scaling normalization and split into training data (70%) and testing data (30%). The model was trained over 150 epochs with a batch size of 32. Evaluation results showed a Mean Absolute Error (MAE) of 0.28 and a Root Mean Squared Error (RMSE) of 0.39, indicating high prediction accuracy. The distribution of prediction results confirmed previous studies indicating that earthquakes in Southern part of Sumatra frequently occur in Bengkulu, western South Sumatra, and Southwestern Lampung. These findings underscore the importance of ongoing seismic hazard mitigation efforts and sustainable development planning in earthquake-prone areas.]]>
Copyrights © 2026
