Build a machine learning-powered web application that predicts flight delays using GitHub Copilot as your AI pair programmer.
An end-to-end application where users can:
- Select a day of the week and destination airport
- Get the probability of their flight being delayed by more than 15 minutes
- See color-coded risk levels in a modern web interface
| Layer | Technology |
|---|---|
| Data Science | Python, Pandas, Scikit-learn |
| Backend | Flask REST API |
| Frontend | HTML, CSS, JavaScript |
| AI Pair Programmer | GitHub Copilot |
- Visual Studio Code with the following extensions:
- GitHub Copilot (required)
- Python Extension
- Jupyter Extension
- Python 3.9+ with pip
- Git for cloning the repository
The data/flights.csv contains 271,000+ FAA flight records from 2013 with delay information, carrier codes, and airport details. Perfect for binary classification.
| Chapter | Focus | Skills |
|---|---|---|
| 0. Setup | Environment & Copilot | Configuration |
| 1. ML Model | Data exploration & training | Jupyter, Copilot Chat |
| 2. REST API | Flask endpoints | Backend development |
| 3. Frontend | Web interface | HTML/CSS/JS |
2-4 hours depending on your experience level.
Happy hacking! May your flights be on time!