Yavas Cemil Emre, Chen Lei, Kadlec Christopher, Ji Yiming
Department of Information Technology, Georgia Southern University, Statesboro, GA, USA.
Sci Rep. 2024 Oct 18;14(1):24440. doi: 10.1038/s41598-024-76483-x.
This research breaks new ground in earthquake prediction for Los Angeles, California, by leveraging advanced machine learning and neural network models. We meticulously constructed a comprehensive feature matrix to maximize predictive accuracy. By synthesizing existing research and integrating novel predictive features, we developed a robust subset capable of estimating the maximum potential earthquake magnitude. Our standout achievement is the creation of a feature set that, when applied with the Random Forest machine learning model, achieves a high accuracy in predicting the maximum earthquake category within the next 30 days. Among sixteen evaluated machine learning algorithms, Random Forest proved to be the most effective. Our findings underscore the transformative potential of machine learning and neural networks in enhancing earthquake prediction accuracy, offering significant advancements in seismic risk management and preparedness for Los Angeles.
这项研究通过利用先进的机器学习和神经网络模型，在加利福尼亚州洛杉矶的地震预测方面开辟了新领域。我们精心构建了一个全面的特征矩阵，以最大限度地提高预测准确性。通过综合现有研究并整合新颖的预测特征，我们开发了一个强大的子集，能够估计潜在的最大地震震级。我们的突出成就是创建了一个特征集，当与随机森林机器学习模型一起应用时，在预测未来30天内的最大地震类别方面具有很高的准确性。在评估的16种机器学习算法中，随机森林被证明是最有效的。我们的研究结果强调了机器学习和神经网络在提高地震预测准确性方面的变革潜力，为洛杉矶的地震风险管理和准备工作带来了重大进展。
