Modeling air quality prediction using a deep learning approach: Method optimization and evaluation
Abstract
A temporal sliding model is proposed for PM2.5 concentration long-term predictions. Integrating the optimal time lag of the spatiotemporal correlations can improve model performance. The proposed model achieves higher prediction in the longer time series. The proposed model has strong practicability in making atmospheric management decisions.
- Publication:
-
Sustainable Cities and Society
- Pub Date:
- February 2021
- DOI:
- 10.1016/j.scs.2020.102567
- Bibcode:
- 2021SusCS..6502567M
- Keywords:
-
- Air pollutant;
- Air quality prediction;
- Deep learning;
- Long-term prediction;
- Temporal sliding