pollutiontracker

Air Quality Dashboard For Greensboro, NC

Air quality forecasting service that predicts PM 2.5 (particulate matter) levels for the next 7 days. It gets updated once every day. Regression model is trained on historical air quality and weather data and is monitoted against predicted and real data.

The air quality prediction

Forecast

Model Performance Monitoring

Predictions vs Outcomes

Hindcast

Architecture

The system follows a feature store based architecture:

  1. Feature history Pipeline: Loads historical air quality and weather data
  2. Feature Pipeline: Computes and stores features in the feature store
  3. Training Pipeline: Trains XGBoost model using historical features
  4. Inference Pipeline: Generates 7-day forecasts using weather forecast data
  5. Dashboard: Displays predictions and historical trends