FORMULA 1 DATA PROJECTS

Hobby project | Data analysis and machine learning | 2026 (ongoing)

A collection of Formula 1 data projects built around the FastF1 API, combining a season-wide exploratory analysis of the 2026 championship with a machine learning model that predicts the Belgian Grand Prix winner from qualifying data. As a Formula 1 fan this is my favourite kind of dataset: complex, unpredictable and always growing.

Key contributions:

Collected qualifying sessions (2021 to 2026) and race results (2021 to 2025) for the Belgian GP through the FastF1 API and engineered features such as best qualifying time and improvement gaps between Q1, Q2 and Q3.

Trained and compared Logistic Regression and Random Forest classifiers for the race winner; Logistic Regression reached a ROC-AUC of 0.935 and correctly picked the 2026 Belgian GP winner.

Performed exploratory analysis of the 2026 season covering driver and constructor performance, points, podiums and DNFs, with interactive Plotly charts.

Produced a race-weekend deep dive into the Monaco 2026 Grand Prix, analysing both qualifying and the race.

Outlined next steps, including weather features and safety car likelihood, to move beyond qualifying-only predictions.

Collaborators: Gabriela Słomiany

Software Used: Jupyter Notebook, FastF1, pandas, scikit-learn, Plotly, Matplotlib

Programming Languages: Python

Repository: github.com/imgabrielas/F1-Enthusiast

Project Date: August 2026