Analysis of Bangladesh Earthquake Data Using Data Mining Techniques

Abstract

Abstract Every year many people perish and abate their living due to earthquakes. In this research, former earthquake data between 1918 and 2018 has been collected for the region, Bangladesh which is at 24°00' N and 90°00' E according to latitude and longitude, respectively. These data have been used as a parameter to determine the relationships among several merits of the data set. Therefore, a linear regression algorithm has been exercised for time series forecasting, and the process gives almost a 75-80% success rate in predicting the earthquake zone along with its magnitude. In addition, a statistical map of earthquake magnitudes has been established from the trained data using Weka Explorer, and the evaluation model depends on the clustered instances. The whole analysis has been implemented to develop a website so that people can be aware of earthquakes before the incident. Finally, the place, Snamganj (25°04' 16'' N and 91°24' 13'' E) has been predicted as the riskiest region in Bangladesh for earthquakes on the Sylhet fault zone.