Neural networks put up against the 2024 Formula 1 season: who finishes where on a given Sunday, and what the championship table looks like by the end.
Trained on historical performance data, the models call the 2024 season race by race: each driver's finishing position on a given Sunday, and the championship standings that fall out of those results. Several modelling approaches are lined up side by side to see which one reads a race weekend best.
Tech stack: Python, TensorFlow/Keras, Jupyter
A project by Manuel Cattoni.