A Bayesian Reasoning for Earthquake Prediction Based on IoT System

Abstract

Yogyakarta, one of the cities in Indonesia, has the most active volcano in Indonesia, namely Mount Merapi. Volcanic activity of Merapi often results in earthquakes, either weak or strong earthquakes. Those earthquakes sometimes are dangerous to the community; therefore, early warnings for earthquakes are very necessary. This paper presents development of earthquake prediction system based on internet of things (IoT) that equipped with a well known data mining techniques for prediction called Bayesian reasoning. A sequential data about earthquakes occurrence has been used to predict earthquake occurrence in the future. A cloud database is used to store real time earthquake occurrence data and a mobile application has developed to give early warning and visualise the earthquake data to the users. Testing of the system shows the proposed earthquake prediction system has high accuracy with Mean Squared Error (MSE) value is 0.127.