Richter's Predictor: Modeling Earthquake Damage

On April 25, 2015, a devastating earthquake struck Nepal, causing widespread damage and loss of life. The 7.8 magnitude earthquake, along with several powerful aftershocks, destroyed or damaged thousands of buildings, homes, and infrastructure, leaving many people homeless and without basic necessities. The problem of predicting earthquake damage to buildings is a critical one, as it can help emergency responders and city planners make informed decisions in the aftermath of a disaster. In fact, a massive effort was undertaken to collect data on the affected buildings and structures. This data was collected and compiled into an extensive dataset, which has been used for a variety of research and analysis purposes. One important use of this dataset has been to train machine learning algorithms to predict the damage caused by earthquakes. The goal of this report is to provide a detailed analysis of these algorithms, a discussion of the results and potential applications of the models developed. In order to develop a robust machine learning algorithm, some basics preprocessing steps have been applied.