Predicting concentration levels of air pollutants by transfer learning and recurrent neural network
作者：
Highlights：
• Accurate prediction for air pollutant concentration for stations with fewer data.
• The stations have even missing data and multiple nearby station data are processed.
• Transfer learning to initialize conveniently recurrent neural networks.
• Empirical analysis on a real sample: more than 12 years for one day ahead prediction.
• Values from air pollutants as well as classical meteorological data are computed.
摘要
•Accurate prediction for air pollutant concentration for stations with fewer data.•The stations have even missing data and multiple nearby station data are processed.•Transfer learning to initialize conveniently recurrent neural networks.•Empirical analysis on a real sample: more than 12 years for one day ahead prediction.•Values from air pollutants as well as classical meteorological data are computed.
论文关键词：Forecasting,Environment monitoring,Transfer learning,Recurrent neural network,Airborne particle matter
论文评审过程：Received 11 March 2019, Revised 30 January 2020, Accepted 5 February 2020, Available online 8 February 2020, Version of Record 24 February 2020.
论文官网地址：https://doi.org/10.1016/j.knosys.2020.105622
