New Publication from SACAQM Team

A new research paper has been published titled:

“Transfer Learning from Wi-Fi Access Point Data to Air Quality Monitoring: A Spatial-Temporal Graph-Based Approach.”

The study presents a novel framework that applies transfer learning to air quality monitoring by leveraging knowledge from Wi-Fi access point data. Using a combination of Graph Neural Networks (GNNs) and temporal models, the approach captures complex spatial-temporal patterns and transfers learned representations to improve air quality forecasting.

Results show that the proposed method achieves accuracy comparable to baseline models while significantly reducing training time and computational cost. Notably, the framework reduces training epochs by over 50% and cuts computation time by more than half.

The approach also enables efficient adaptation to new sensor deployments without requiring full retraining, highlighting its relevance for resource-constrained and scalable air quality monitoring systems.

🔗 Read the full paper here: https://ieeexplore.ieee.org/document/11472283