Tremor Track: A Machine Learning Based Earthquake Prediction System

Earthquakes are the most destructive natural catastrophe because they cause widespread destruction of life, property, and infrastructure, especially in the densely populated and seismically active regions. This paper describes developing and implementing a real-time earthquake prediction system called “Tremor Track”. It basically looks at developing a machine learning model based on the Random Forest algorithm that predicts how likely an earthquake is with historical records and geophysical data for better disaster preparedness and risk migration. With the system, any place in the world may have its earthquake proneness for a user to predict and quantify just by mentioning three parameters on the desired location that are latitude, longitude, and elevation from sea level. The Experimental work results show that “Tremor Track” is capable of effectively tracking the patterns and correlations in seismic data, as a result, making more accurate predictions of earthquakes with higher success rate reliability. Through the algorithm of Random Forest, “Tremor Track” combines historical seismic events, latitude, longitude, and geophysical properties to decide earthquake likelihood in certain regions. “Tremor Track” also supplies an estimated outcome: how prone the given city is to earthquakes is determined by it.