Madhushree N * and Thiruukumaran
Earthquake is one of the most hazardous, devasting natural calamity and yet a very least predictable natural disaster that occur. Prediction of earthquake has been a challenging research for many researchers. With the increasing amount of earthquake dataset collected, many researchers try to solve the task of predicting the earthquake in future time. Even though many data mining techniques are been used, the prediction rate is not still accurate due to lack of feature extraction technique. The proposed methodology enhances the performance of earthquake prediction. As obtained precursory pattern features CART algorithm is used to get accurate prediction of both the magnitude range and an effective time range of future earthquakes.