📝 Description

The project focuses on applying advanced deep learning techniques to forecast spatiotemporal weather patterns and events. The student will work with the data from the MET Norway weather API, which provides high-resolution weather forecasts and historical observations. Once the data is prepared and pre-processed by building a spatiotemporal graph representation, the student will implement and fine-tune Spatiotemporal Graph Neural Networks (STGNNs) models using PyTorch and Torch Spatiotemporal, experimenting with different architectures and hyperparameters. By the end of the project, the student will have developed a strong understanding of meteorological data pre-processing, graph-based modeling, spatiotemporal deep learning, and its application to real-world weather forecasting tasks, which can be further explored in future research, as well as potential experience in deploying these models in production environments.

Data: The student will use raw high-resolution weather data from the MET Norway API, which includes various meteorological variables such as temperature, precipitation, and wind speed.

📨 Contact:

Filippo Maria Bianchi filippo.m.bianchi@uit.no Roberto Neglia roberto.neglia@uit.no