Earthquake Prediction using Machine Learning
Earthquake prediction is a complex and critical problem that has significant implications for human safety and infrastructure. Although the exact prediction of earthquakes remains challenging due to the non-linear nature of seismic data, machine learning approaches can be used to identify patterns and predict the likelihood of seismic events based on historical data. This report explores the use of machine learning techniques for earthquake prediction by analyzing seismic data and building predictive models.
Dataset
The National Earthquake Information Center (NEIC) determines the location and size of all significant earthquakes that occur worldwide and disseminates this information immediately to national and international agencies, scientists, critical facilities, and the general public. The NEIC compiles and provides to scientists and to the public an extensive seismic database that serves as a foundation for scientific research through the operation of modern digital national and global seismograph networks and cooperative international agreements. The NEIC is the national data center and archive for earthquake information. This dataset includes a record of the date, time, location, depth, magnitude, and source of every earthquake with a reported magnitude 5.5 or higher since 1965.
Results
Model Accuracy (%) Linear Regression 82.5 SVM (Support Vector Machine) 83.0 Decision Tree 90.5 MLP (Multilayer Perceptron) 90.5 Based on this table, Decision Tree and MLP both have the highest accuracy at 90.5%, while Linear Regression has the lowest accuracy of 82.5%.
