A Comparative Study of Machine Learning Techniques for Seismic Activity Monitoring using Fiber Optic Distributed Acoustic Sensors

Optical fiber based Distributed Acoustic Sensor records the seismic waves from earthquakes and can be employed to detect the presence of earthquakes. Since these sensors generate large amount of data, it becomes imperative to use artificial intelligence techniques for deriving meaningful insights. In this work, the effectiveness of different machine learning models in detecting seismic events is investigated. As such, the models are trained and tested using data acquired from a public database. Among the various models investigated, Random Forest Classifier shows a better accuracy of 88.2% compared to other models.