Derin Öğrenme Yöntemleri Kullanılarak Deprem Tahmini Gerçekleştirilmesi
The exact prediction of major earthquakes is one of the important issues. The purpose of the forecast is to take the necessary measures to reduce deaths and destruction by knowing about earthquakes in advance. The study and analysis of earthquakes is very important for earthquake prediction. A study that can accurately predict earthquakes has not yet taken place. However, researchers are working to create some earthquake patterns by studying the records of previous earthquakes. Various studies are carried out to predict future earthquakes. These studies have focused on statistical methods in general, and more recently on artificial intelligence and deep learning. Deep learning methods, one of the fields of artificial intelligence, are used to make realistic predictions. Recurrent Neural Network (RNN), one of the deep learning architectures, is preferred because of its high success rate in solving problems such as prediction and classification. In this study, a model has been proposed for estimating the magnitude of a possible earthquake by using information such as date, latitude, longitude and depth of the earthquake from earthquake data in Turkey. In addition, the Long-Short-Time Memory (LSTM) model, which is a type of RNN, is used to predict the time together with the magnitude of the earthquake that will occur. The proposed model has been trained with the time series in the prepared earthquake data set and estimation has been made.
The proposed model estimated the magnitudes of the Kalaba-Sivrice (Elazıg) earthquake (24.10.2020) and the Aegean Sea-(Izmir) (30.10.2020) earthquake with 91% and 93% probabilities, respectively.
Copyright (c) 2022 Artificial Intelligence Studies
This work is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License.
Artificial Intelligence Studies (AIS) publishes open access articles under a Creative Commons Attribution 4.0 International License (CC BY). This license permits user to freely share (copy, distribute and transmit) and adapt the contribution including for commercial purposes, as long as the author is properly attributed.
For all licenses mentioned above, authors can retain copyright and all publication rights without restriction.