A Spatio-Temporal Method for Air Quality Index Prediction with Diverse Sensor Quality Types

Air pollution critically affects public health, requiring accurate spatial and temporal air quality predictions. This study predicts the Air Quality Index (AQI) in Delhi, India, using high-cost and simulated low-cost sensors in Wireless Sensor Networks (WSNs). Temporal prediction utilizes Support Vector Regression (SVR), while spatial prediction uses a Deep Learning Neural Network (DLNN). The findings represent low-cost sensors as a cost-effective alternative for air quality monitoring and accurate AQI predictions.