Journals / International Journal of Pure and Applied Sciences / 2026 / Volume: 12 Issue: 1

A Comprehensive Benchmarking of Classical and Deep Learning Models for Global Earthquake Magnitude Prediction

Çiğdem Ceylan(MALATYA TURGUT OZAL UNIVERSITY)0000-0002-5961-3085
Serpil Aslan(MALATYA TURGUT ÖZAL ÜNİVERSİTESİ)0000-0001-8009-063X
Pages
—
Publication Date
2026-06-22
🏛️Bibliographic metadata and full text access for this article are provided via DergiPark.
DergiPark Record ↗
✨

AI Summary & Key Takeaways

AI Insight
⚡ TL;DR(Too Long; Didn't Read)

Findings indicate that Random Forest significantly outperforms all other models, delivering the lowest error rates and highest explanatory power.

📌 3 Key Takeaways

🎯
Objective & Hypothesis:Earthquakes cause severe loss of life and widespread destruction, making accurate magnitude estimation a critical component of disaster management.
📊
Key Findings & Data:Furthermore, only a limited number of studies evaluate model performance using multiple and diverse error metrics, highlighting the need for a more systematic and unified analytical framework.
💡
Conclusion & Impact:Findings indicate that Random Forest significantly outperforms all other models, delivering the lowest error rates and highest explanatory power.
🔬Design:Modeling & Simulation
📍Scope:2023 Verileri
⚙️Variables:Earthquake Magnitude Estimation · Machine Learning · Deep Learning

Abstract

Earthquakes cause severe loss of life and widespread destruction, making accurate magnitude estimation a critical component of disaster management. Although literature contains numerous studies, many rely on a limited set of models or specific data types, thereby failing to provide a comprehensive and holistic comparison between classical machine learning and deep learning approaches. Furthermore, only a limited number of studies evaluate model performance using multiple and diverse error metrics, highlighting the need for a more systematic and unified analytical framework. This study presents a systematic and comprehensive comparative analysis of classical machine learning and deep learning models for earthquake magnitude estimation using a global seismic dataset covering the period from 1995 to 2023. Classical approaches such as Linear Regression, Decision Trees, Random Forest, K-Nearest Neighbor, and Support Vector Regression were evaluated alongside advanced deep learning architectures, including MLP, CNN, LSTM, BiLSTM, CNN-LSTM, CNN-GRU, and GRU. Model performances were assessed using a set of widely adopted regression-based evaluation metrics, including MAE, MSE, RMSE, R², MSLE, and Log-Cosh, supported by detailed prediction–actual comparisons, residual error distribution analyses, and time-series visualizations. Findings indicate that Random Forest significantly outperforms all other models, delivering the lowest error rates and highest explanatory power. While deep learning models offer meaningful predictions in certain contexts, they remain less effective due to the characteristics of the dataset. Overall, the study provides a rigorous benchmark and valuable reference for data-driven risk analysis, early warning systems, and machine learning-based disaster management applications.

🔖

Cite This Article

Tanyıldız, H., Ceylan, Ç., Aslan, S. (2026). A Comprehensive Benchmarking of Classical and Deep Learning Models for Global Earthquake Magnitude Prediction. International Journal of Pure and Applied Sciences, Volume: 12 Issue: 1. https://doi.org/10.29132/ijpas.1838407

📄 Full Text PDF Viewer

📄

PDF görüntüleyici hazırlanıyor...

References (25)

  1. A. Sebatlı Sağlam and F. Çavdur, “Yapay sinir ağı ile deprem şiddeti tahmini: Farklı ağ tasarımlarının ve eğitim algoritmalarının incelenmesi,” Gazi Üniversitesi Mühendislik Mimarlık Fakültesi Dergisi, vol. 37, no. 4, pp. 2133–2146, 2022, doi: 10.17341/gazimmfd.791337.
  2. G.-A. Tselentis and J. Zahradnik, “Aftershock modelling based on empirical and statistical methods,” Soil Dynamics and Earthquake Engineering, vol. 24, no. 11, pp. 849–857, 2004.
  3. C. H. Yeh, Y. L. Lin, and C. C. Hsu, “Earthquake prediction model using Random Forest,” Geophysical Journal International, vol. 217, no. 1, pp. 89–100, 2019, doi: 10.1093/gji/ggz011.
  4. M. Asim, M. S. Siddiqui, and M. Ahmad, “Earthquake magnitude prediction using recurrent neural networks,” Journal of Seismology and Earthquake Engineering, vol. 22, no. 3, pp. 45–56, 2020.
  5. M. Hosseini, A. Azizi, and G. G. Amiri, “Hybrid deep learning model for earthquake magnitude prediction using CNN-LSTM,” Natural Hazards Review, vol. 22, no. 4, Art. no. 04021035, 2021, doi: 10.1061/(ASCE)NH.1527-6996.0000460.
  6. L. Wang, Y. Chen, and X. Liu, “Gated recurrent unit model for real-time seismic data analysis,” Seismological Research Letters, vol. 94, no. 1, pp. 12–21, 2023, doi: 10.1785/0220220183.
  7. Y. Zhang, K. Tang, and J. Li, “A transformer-based approach for seismic anomaly detection in earthquake forecasting,” IEEE Access, vol. 11, pp. 45678–45689, 2023, doi: 10.1109/ACCESS.2023.3267894.
  8. X. Zhu, R. Guo, and D. Sun, “Challenges and opportunities in earthquake prediction with deep learning,” Computers & Geosciences, vol. 151, Art. no. 104752, 2021, doi: 10.1016/j.cageo.2021.104752.
  9. M. Alimohammadi and A. R. Khaloo, “Ensemble learning-based approach for rapid earthquake damage assessment using remote sensing data,” Remote Sensing, vol. 14, no. 9, Art. no. 1982, 2022, doi: 10.3390/rs14091982.
  10. G.-A. Tselentis and J. Zahradnik, “Aftershock modeling based on empirical and statistical methods,” Soil Dynamics and Earthquake Engineering, vol. 24, no. 11, pp. 849–857, 2004, doi: 10.1016/j.soildyn.2004.06.008.
  11. H. Liu, X. Zhang, and Y. Gao, “Machine learning techniques for earthquake magnitude prediction: A comparative study,” Seismological Research Letters, vol. 93, no. 5, pp. 2781–2793, 2022, doi: 10.1785/0220220045.
  12. I. Tsamardinos and C. F. Aliferis, “Earthquake magnitude prediction using support vector machines and hybrid kernel optimization,” Engineering Applications of Artificial Intelligence, vol. 92, Art. no. 103632, 2020, doi: 10.1016/j.engappai.2020.103632.
  13. Y. Hu, H. Liang, and R. Wang, “A convolutional neural network-based model for earthquake magnitude prediction using seismic waveform data,” Applied Sciences, vol. 11, no. 4, Art. no. 1765, 2021, doi: 10.3390/app11041765.
  14. S. Hochreiter and J. Schmidhuber, “Long short-term memory,” Neural Computation, vol. 9, no. 8, pp. 1735–1780, 1997, doi: 10.1162/neco.1997.9.8.1735.
  15. I. G. A. Kusuma and H. A. Nugroho, “Bidirectional LSTM for earthquake magnitude prediction using seismic time series data,” Procedia Computer Science, vol. 197, pp. 54–62, 2022, doi: 10.1016/j.procs.2021.12.062.
  16. M. Rashid and M. A. Shah, “Deep hybrid models for earthquake prediction: A CNN-GRU approach,” Geoscience Frontiers, vol. 13, no. 5, pp. 1440–1452, 2022, doi: 10.1016/j.gsf.2021.101247.
  17. S. Aslan, “BiGRU-CNN tabanlı derin öğrenme modeliyle Türkiye’deki COVID-19 aşılarına yönelik Twitter duygu analizi,” International Journal of Pure and Applied Sciences, vol. 8, no. 2, pp. 312–330, 2022.
  18. Warcoder, “Earthquake Dataset,” Kaggle, 2023. [Online]. Available: https://www.kaggle.com/datasets/warcoder/earthquake-dataset
  19. G. James, D. Witten, T. Hastie ve R. Tibshirani, An Introduction to Statistical Learning: With Applications in R, 2. Baskı. New York, NY, ABD: Springer, 2021.
  20. M. Khedimi, T. Zhang, C. Dehmani, X. Zhao, and Y. Geng, “A unified deep learning ensemble framework for voice-based Parkinson’s disease detection and motor severity prediction,” Bioengineering, vol. 12, no. 7, Art. no. 699, 2025.
  21. S. M. Robeson and C. J. Willmott, “Decomposition of the mean absolute error (MAE) into systematic and unsystematic components,” PLOS ONE, vol. 18, no. 2, Art. no. e0279774, 2023.
  22. T. Chai and R. R. Draxler, “Root mean square error (RMSE) or mean absolute error (MAE)?—Arguments against avoiding RMSE in the literature,” Geoscientific Model Development, vol. 7, no. 3, pp. 1247–1250, 2014.
  23. D. Chicco, M. J. Warrens, and G. Jurman, “The coefficient of determination R-squared is more informative than SMAPE, MAE, MAPE, MSE and RMSE in regression analysis evaluation,” PeerJ Computer Science, vol. 7, Art. no. e623, 2021.
  24. M. R. Salmanpour, M. Alizadeh, G. Mousavi, S. Sadeghi, S. Amiri, M. Oveisi, and I. Hacihaliloglu, “Machine learning evaluation metric discrepancies across programming languages and their components in medical imaging domains: Need for standardization,” IEEE Access, 2025.
  25. A. Botchkarev, “Performance metrics (error measures) in machine learning regression, forecasting and prognostics: Properties and typology,” arXiv preprint, arXiv:1809.03006, 2018.