Sign in Subscribe
Machine learning

Real-Time Air Pollution Monitoring in Delhi Using Online Anomaly Detection Algorithms for Streaming Data

Strategic Insights

— 1 min read

A recent study by researchers from Delhi Technological University has found that online machine learning algorithms can be effectively used for real-time air quality monitoring in Delhi, India. The study, which utilized nine popular unsupervised online anomaly detection models, demonstrated that these algorithms can detect anomalies in air quality data streams with high accuracy. The research focused on monitoring gaseous pollutants in real-time at 20 base stations corresponding to the most polluted sites in the city between 1st May 2022 and 30th April 2023.

Key Takeaways:

  • The study utilized nine popular unsupervised online anomaly detection models, including Half Space Trees (HST), Quantile Filter One-Class Support Vector Machine (QF_OCSVM), and Quantile Filter Half Space Trees (QF_HST).
  • The performance of each model was evaluated using four metrics: Accuracy, ROCAUC, Recall, and Geometric Mean.
  • The QF_OCSVM, QF_HST, and HBOS models achieved the highest scores across four performance metrics.
  • The study demonstrated the effectiveness of online machine learning algorithms for real-time air quality monitoring in Delhi.
  • The research highlighted the importance of tracking pollution levels in cities, as hazardous pollution can seriously affect human health.
  • The study's findings can be applied to other cities with similar pollution challenges.

Statistics:

  • 9 unsupervised online anomaly detection models were used in the study.
  • 20 base stations were used to monitor gaseous pollutants in real-time.
  • The duration of the study was from 1st May 2022 to 30th April 2023.
  • The performance of each model was evaluated using four metrics: Accuracy, ROCAUC, Recall, and Geometric Mean.
  • The QF_OCSVM, QF_HST, and HBOS models achieved the highest scores across four performance metrics with accuracy rates of over 90%.

Sources:

  • NewsRx. Findings from Delhi Technological University Has Provided New Data on Data Streaming (Real-time Air Pollution Monitoring In Delhi Using Online Anomaly Detection Algorithms for Streaming Data). Ecology, Environment & Conservation. August 15, 2025; p 170.
  • Real-time Air Pollution Monitoring In Delhi Using Online Anomaly Detection Algorithms for Streaming Data. National Academy Science Letters, 2025.

* National Academy Science Letters can be contacted at: Springer India, 7TH Floor, Vijaya Building, 17, Barakhamba Road, New Delhi, 110 001, India.

* Springer can be contacted at: www.springer.com

* National Academy Science Letters can be contacted at: www.springerlink.com/content/0250-541x/

Read more