Reliable earthquake forecasting methods have long been sought after, and so the rise of modern data science techniques raises a new question: does deep learning have the potential to learn this pattern? 
In this study, we leverage the large amount of earthquakes reported via good seismic station coverage in the subduction zone of Japan.  We pose earthquake forecasting as a classification problem and train a Deep Learning Network to decide, whether a timeseries of length ≥ 2 years will end in an earthquake on the following day with magnitude ≥ 5 or not.
 
Our method is based on spatiotemporal b value data, on which we train an autoencoder to learn the normal seismic behaviour. We then take the pixel by pixel reconstruction error as input for a Convolutional Dilated Network classifier, whose model output could serve for earthquake forecasting. We develop a special progressive training method for this model to mimic real life use. The trained network is then evaluated over the actual dataseries of Japan from 2002 to 2020 to simulate a real life application scenario. The overall accuracy of the model is 72.3%. The accuracy of this classification is significantly above the baseline and can likely be improved with more data in the future.