Air Quality Monitoring Network: Current Measurement Data Vienna

Air Quality Monitoring Network: Current Measurement Data Vienna

This deposit contains all research outputs from a machine learning experiment on urban air quality prediction, conducted as part of the Data Stewardship (DaSt 2026) — FAIR Data Science course at TU Wien. The project uses sensor data from 17 monitoring stations across Vienna (source: MA22 / data.gv.at) to predict PM10 fine particulate matter concentrations using Random Forest and XGBoost models.

Included files: processed training, validation, and test datasets (CSV); trained model files (Random Forest .pkl, XGBoost JSON); prediction output CSV; visualisation charts (histograms, performance comparison, confusion matrix, PNG); Python source code with README and requirements.txt; and this Data Management Plan (PDF). Raw input data is not archived here as it remains permanently available at data.gv.at.

Licence: Data and outputs — CC BY 4.0; Source code — MIT Licence.

Keywords: air quality, PM10, Vienna, machine learning, Random Forest, XGBoost, open government data, FAIR data, environmental monitoring

Related resource: https://www.data.gv.at (original sensor data, OGD Austria / MA22)