Forecasting earthquakes is challenging due to the difficulty in detecting underground phenomena. However, statistically, 30% of large earthquake foreshocks occur before the main shock, which indicates that, theoretically, it is possible to forecast them. However, systematically collecting historical earthquake data has been challenging because of the lack of computational data collection and procedures. Addressing these limitations in historical data collection, we have compiled a database of earthquake events in Türkiye from 2004 to 2023, encompassing 187,399 cases with magnitudes ranging from 2.0 ML to 7.0 ML. This database includes parameters such as latitude, longitude, date, and magnitude. With the database, we built a neural network model capable of predicting earthquakes over 2.0 ML across Turkey up to one month in advance. To evaluate the model, we made it to predict earthquakes in January 2024 and matched the result to the actual events. The pattern between the two was 98% the same. Additionally, the model forecasted an earthquake over 4.0 ML in Istanbul between February and March, which happened on February 19th. Though this prediction model can predict one month's future and various evaluations must be made throughout the time, this paper carries the value of potential.

Keywords - Earthquake Prediction, Neural Network Model, Artificial Intelligence, Machine Learning, Türkiye

[1] Siat Yıldırım, An Observation Study On The Psychosocial Effect of The 6 February Kahramanmaraş Earthquake And The Importance Of Sustainable Response To Earthquake Victims, 2023.

[2] Amirul Hoque, Jyoti Raj , Dr. Ashim Saha. “Approaches of Earthquake Magnitude Prediction using machine learning techniques.”. International Conference on Computational Intelligence & IoT (ICCIIoT) (2018): 1-5.

[3] Wyss, Max. "Cannot earthquakes be predicted?." Science 278.5337 (1997): 487-490.

[4] Braile and Brehm 1998; Knopoff 2000; Kirschvink 2000; Ellsworth et al. (1999)

[5] Turcotte, Donald L. "Earthquake prediction." Annual review of earth and planetary sciences 19.1 (1991): 263-281.

[6] Geller, Robert J., et al. "Earthquakes cannot be predicted." Science 275.5306 (1997): 1616-1616.

[7] Anderson, J.G., 1981. A simple way to look at a Bayesian model for the statistics of earthquake prediction. Bulletin of the Seismological Society of America, 71(6), pp.1929-1931.

[8] Varotsos P. and Alexopoulos K., 1984. Physical properties of the variations of the electric field of the earth preceding earthquakes, I. Tectonophysics, 110(1-2), pp.73-98

[9] Wang W., Cao X. and Song X., 2001. Estimation of the Earthquakes in Chinese Mainland by using artificial neural networks. Chinese Journal of Earthquakes, 3(21), pp.10-14.

[10] Reyes, Jorge, Antonio Morales-Esteban, and Francisco Martínez-Álvarez. "Neural networks to predict earthquakes in Chile." Applied Soft Computing 13.2 (2013): 1314-1328.

[11] Wang, Xi, et al. "Small Earthquakes Can Help Predict Large Earthquakes: A Machine Learning Perspective." Applied Sciences 13.11 (2023): 6424.

[12] Manral, Gaurav Singh, and Alka Chaudhary. "Prediction of Earthquake Using Machine Learning Algorithms." 2023 4th International Conference on Intelligent Engineering and Management (ICIEM). IEEE, 2023.

[13] Handan Çam, Osman Duman, “Yapay Sinir Ağı Yöntemiyle Deprem Tahmini: Türkiye Batı Anadolu Fay Hattı Uygulaması”,2016

[14] https://www.afad.gov.tr

[15] Liu, M.; Stein, S. Mid-continental earthquakes: Spatiotemporal occurrences, causes, and hazards. Earth-Sci. Rev. 2016, 162, 364–386.

[16] G. P. Zhang, "Neural networks for classification: a survey.", Systems, Man, and Cybernetics, Part C: Applications and Reviews, IEEE Transactions on, vol. 30, pp. 451-462, 2000.

[17] Isaac Martín De Diego, Ana R. Redondo, Rubén R. Fernández, Jorge Navarro, Javier M. Moguerza. “General Performance Score for classification problems”. 2022