Earthquake Magnitude Prediction Using Machine Learning Technique

The prophecy of an earthquake is a very complicated task which has been resolved in different ways. However, there is a need for an elementary definition along with adequate catalogues. These areas have been marked with widely used geological parameters. Total characteristics of study are based on not only the training and testing data sets but also qualitative data about earthquake prediction. The machine learning technique is applied on Japan, Turkey, Greece and Indian Subcontinent regions. The model show relationship be-tween calculated seismic data and future earthquake occurrences. It is the setup which has been fixed to provide the forecasts for earthquakes of a magnitude of 5.0 and its above for Japan, 4.5 and greater for Indian Subcontinent region, 3.7 and greater for Turkey region, 3.2 and greater for Greece region, before fifteen days to the earthquakes. A noticeable improvement has been produced by the model in earthquake predictions and compared to previous prediction studies. The studies bestowed during this work analysing the performance of the prediction across the earthquake magnitude statistics of the target seismogenic zones which will help in the further development and improvement of the system.