Causal Spatiotemporal Graph Neural Network

A Causal Spatiotemporal Graph Neural Network for Interpretable PM2.5 Forecasting in Delhi NCR presents an advanced deep learning framework that combines causal inference, spatial relationships, and temporal dependencies to accurately predict PM2.5 air pollution levels in the Delhi NCR region. The book explores how Graph Neural Networks (GNNs) can model interactions among monitoring stations while identifying causal factors influencing air quality. It emphasizes forecasting accuracy, interpretability, and explainable AI, enabling policymakers and environmental agencies to better understand pollution patterns and make informed decisions for air quality management.