Datathon 2026: Access to a Livable Planet

Using EPA AQI data, we apply machine learning and data visualization to uncover hidden air quality risk patterns across U.S. counties.

Repository Contents

data_cleaning.ipynb
Loads, merges, and cleans raw EPA AQI by county data. Handles missing values, invalid ratios, feature normalization, and constructs derived indicators used across all downstream analyses.
high_risk_cluster.ipynb
Uses K-Means clustering to group counties based on long-term air quality profiles.
Focuses on exposure frequency, pollutant composition, and historical intensity to identify high-risk air quality regimes.
extreme_event_early_warning.ipynb
Develops an XGBoost-based early warning model for future extreme AQI events using historical trends and event recency features.
Emphasizes interpretability and practical risk signals rather than short-term forecasting accuracy.

🌐 Website: https://runa-1117.github.io/datathon-2026/
📁 Google Drive: https://drive.google.com/drive/u/0/folders/1e-E_GUNo2w_4CTxRWZUdz8_r0qKi5bW5
📊 Tableau Public Dashboard 1: https://public.tableau.com/app/profile/anna.huang8759/viz/Datathon2026_17705177818560/Dashboard3#1
📊 Tableau Public Dashboard 2: https://public.tableau.com/app/profile/anna.huang8759/viz/Datathon2026-dashboard2-breast-cancer/Dashboard1

Runa He, Marco Guo, Anna Huang, Amelia Li • DubsTech Datathon • 2026

Name		Name	Last commit message	Last commit date
Latest commit History 14 Commits
.gitignore		.gitignore
Max_AQI.ipynb		Max_AQI.ipynb
README.md		README.md
data_cleaning.ipynb		data_cleaning.ipynb
extreme_event_early_warning.ipynb		extreme_event_early_warning.ipynb
high_risk_cluster.ipynb		high_risk_cluster.ipynb