← projects  /  EPQ F1 Predictive Model
PyTorch · TabNet

EPQ F1 Predictive Model

Year 2025
Stack PyTorch · TabNet · pandas · scikit-learn · Python

Neural networks and TabNet for Formula 1 race finishing position prediction. Full reproducible pipeline — feature engineering, strict leakage prevention, season-by-season backtesting.

Overview

This project builds a predictive model for Formula 1 race finishing positions using structured data pipelines, feature engineering and modern machine learning methods. The aim is to test how far neural networks and TabNet can extract patterns from timing, telemetry and historical race information.

Objectives

Methods

Role

Tech Stack

Python, PyTorch, TabNet, pandas, NumPy, scikit-learn

Results

Metrics, learning curves and predictions will be added here when full evaluation is completed.