Formula 1 Prediction AI

Technologies

Overview

Worked with a classmate to develop a predictive model for determining F1 race winners. Using historical F1 data found on Kaggle, we cleaned and structured relevant data to be used in training the model. Data points such as historical winners, qualifying results, and points standings. We conducted a exploritory analysis on the impact of qualifying positon on finishing position which found that there is a strong correlation between a driver's starting and finishing position. We trained a linear regression, neural network classifier, and neural network regressor. We were able to achieve a 68% score on our neural network regressor when tested against the 2021 F1 season results. If you watch F1, you know how crazy that season was in terms of battling for 1st place.