Initial release

Key Features and Improvements ✨

• Complete Forecasting Pipeline: A robust system that automates the entire process of fetching live Air Quality Index (AQI) data, training a predictive model, and deploying it.
• XGBoost Model: This release includes an XGBoost model variant for forecasting. It utilises recursive forecasting and lagged windows to predict all features, improving the accuracy and comprehensiveness of the predictions.
• Automated Workflows: Integrated GitHub Actions to automate key parts of our development cycle. We now have workflows for:
  • Docker Image Building: Automatically builds a new Docker image whenever source code or scripts are updated, ensuring the deployment environment is always up-to-date.
  • Data Fetching: Automatically retrieves and saves the latest AQI data.
  • Model Training: Automatically trains the model.
  • Model Deployment: Can be triggered to deploy the best model.