Prediction of earthquake based on artificial neural network technique

• Main Authors: Saleem, Abrar Khalid, Rashed, Ahmed Noori
• Format: Conference Proceeding
• Language: English
• Subjects: Artificial intelligence; Artificial neural networks; Earth movements; Earthquake prediction; Earthquakes; Internal energy; Mathematical analysis; Neural networks; Performance evaluation; Plates (tectonics)
• ISSN: 0094-243X; 1551-7616
• DOI: 10.1063/5.0190668

Earthquake are nature’s calamity produced by the movement of the earth’s tectonic plates as a result of its enormous internal energy being released, the earthquake occurrences prediction, help reduced magnitude of destruction minimized. For predicting an earthquake’s time, magnitude, depth and location of the earthquake, a variety of techniques have been suggested, such as statistical and mathematical analysis and a signal investigation of precursors, due to an ostensibly dynamic character of seismic, they usually do not produce excellent results. The capacity of artificial intelligence to detect hidden patterns of data and nonlinear relation well-known has been gaining attention in recent years. This work applied feed forward neural network to predict the next earthquake occurrence based on historical data got data from General Directorate of Meteorology and Seismic Monitoring in Iraq, through study data for three different regions in Iraq (Sulaymaniyah, Maysan and Wasit to predict earthquake characteristics (time, magnitude, location, and depth) and evaluating the performance of the prediction using testing data. The results of experiments show that system present good performance.