Google-Harvard Initiative To Help Predict Earthquake Aftershocks Using Machine Learning

An initiative is being undertaken by scientists from Harvard and Google whereby machine learning would be used to predict the locations of earthquake aftershocks. This initiative is expected to reduce the damage and loss caused to both life and natural property.

It should be noted that earthquakes occur in sequences with a series of aftershocks following the main quake. These aftershocks while not on the same scale as the main quake often cause major damage and impede rescue and relief efforts. Timing and size of aftershocks can be predicted under current technology and it is expected that this machine learning initiative would enable emergency services to be aware of areas at risk of aftershocks. By predicting the possible areas at risk of aftershocks, measures such as evacuation of populace can be put in place by the concerned authorities minimizing possible casualties and damage.

A database of information on around 118 major earthquakes was compiled for the purposes of this exercise which also included the Bhuj earthquake of 2001.