This repository contains a relatively simple deep learning model that would score in the top 4 of the recent Kaggle competition "LANL Earthquake Prediction". This approach does not involve ensembling, extensive feature engineering (single simple feature), or a complex network architecture, and is a fundamentally different approach than the one taken by the other top performers in this competition.
Link to competition for background: https://www.kaggle.com/c/LANL-Earthquake-Prediction/overview
Deep learning model is a CRNN (Convolutional Recurrent Neural Network)