Jerusalem Temperature Prediction Using Deep Learning

Abstract

Due to the recent surge in deep learning and machine learning research aiming to solve previously unsolvable problems, interest naturally spurred regarding the viability of weather prediction through a data-driven-only approach in comparison with the current physics-based approach. This paper will look into the viability of such an approach using deep learning and machine learning techniques for predicting the temperature at the city of Jerusalem. It will test popular architectures for such a problem, utilise observations from multiple stations, and finally produce a model that outperforms one of the professional models currently used for this task and performs similarly to another.