Earthquake Prediction by Using Time Series Analysis

Abstract

Nowadays, with the developing technologies, big data stored has started to be formed. Data mining methods have been developed due to the need to obtain information from these stored data. These methods include clustering, classification, association rule, and time series. In this study, time series analysis, one of the data mining methods, was emphasized. Time series provide predictions about future time by time data. Time series are divided into two: linear and non-linear methods. Linear time series methods estimate by assuming that the series is stationary. Non-linear time series methods predict based on the raw version of the series in the real world. In this study, artificial neural networks (ANNs), one of the non-linear methods of Time Series Analysis, are used. Earthquake data were discussed with ANNs. Estimates were made by analyzing earthquake data.