Classifying Seismic Events: A Machine Learning Approach to Identifying Earthquakes, Explosions and Other Rare Events
Document Type
Conference Proceeding
Publication Date
6-14-2026
Abstract
Seismic event classification can differentiate between natural and human-caused geophysical events in near-real time, which has numerous potential applications. Using a dataset of 30,000 seismic events from the United States Geological Survey (USGS), machine learning (ML) models were developed and evaluated to automate the classification of earthquakes, explosions, or other events. Key features were identified as depth, magnitude, latitude, and longitude. A Random Forest (RF) model achieved the best performance, with 99.8% accuracy, F1-score of 0.97, and recall of 0.95, which performed better than a trivial majority-class model 90.6% accuracy. Classification of explosions or other rare geophysical events posed a challenge for all models because the dataset was severely imbalanced, with 90.6% of the data classified as earthquakes. RF models incorporate an ensemble approach and place higher class weights across rarer events, making them better for classification of minority events. Neural networks (NN) were also explored. They achieved 98.8% accuracy but underperformed in minority class recall compared to the RF model. The NN highlighted the importance of addressing class imbalances to improve performance, so Synthetic Minority Oversampling (SMOTE) was applied to assist the NN achieve better predictions of rarer seismic events. The classical ML and NN results demonstrate the utility of these models to enhance capabilities by enabling accurate, real-time seismic event classification, potentially contributing to treaty compliance, battlefield awareness, and disaster response applications.
Source Publication
Emerging Trends in Scientific Computing and Theoretical Computer Science
Recommended Citation
Weed, C.R., Wagner, T., Langhals, B., Auclair, P. (2026). Classifying Seismic Events: A Machine Learning Approach to Identifying Earthquakes, Explosions and Other Rare Events. In: Arabnia, H.R., Hodson, D.D., Grimaila, M.R., Wagner, T.J., Maurer, P.M., Deligiannidis, L. (eds) Emerging Trends in Scientific Computing and Theoretical Computer Science. CSCE 2025. Communications in Computer and Information Science, vol 2936. Springer, Cham. https://doi.org/10.1007/978-3-032-22211-4_1
Comments
© The Authors, under exclusive license to Springer Nature Switzerland AG 2026.
The full paper is available from the publisher via subscription or purchase using the DOI link below.
Event: 23rd International Conference, CSC 2025, and 21st International Conference, FCS 2025, Held as Part of the World Congress in Computer Science, Computer Engineering, and Applied Computing, CSCE 2025, Las Vegas, July 21-24, 2025.