Interpretable PM2.5 Forecasting for Urban Air Quality: A Comparative Study of Adaptive Time-Series Models
The study compared lightweight and interpretable forecasting models for hourly PM2.5 prediction in Beijing using time-series data. It tested three models—SARIMAX, Facebook Prophet, and NeuralProphet—under two adaptive deployment strategies. Findings show that Facebook Prophet offers the best balance of accuracy and speed across both methods, proving that transparent models are practical for real-world air quality monitoring.