This project combines statistical research with a web application to explore the effectiveness of linear regression models for weather prediction.
- Analyzed 34 years of historical weather data (1990–2024) from Atlanta Hartsfield-Jackson Airport.
- Applied linear regression and feature engineering to model daily temperatures.
- Conducted error analysis and model evaluation, achieving R² ≈ 0.90 with an average error of ~3°F.
- Reported findings in a 54-page technical paper applying the engineering design process, including evaluation of accuracy, limitations, and potential use cases in public safety, agriculture, and energy.
- Developed a Python-based forecasting tool with a simple, interactive interface.
- Users can upload historical weather data and generate future forecasts using the trained regression models.
- Provides visualizations of predicted vs. actual trends, making model performance easy to interpret.
- Demonstrates how research outcomes can be translated into a practical, user-friendly tool for decision-making.
For a detailed explanation of the methods, results, and analysis, see the full report:
👉Full Research Paper (PDF)