Air Pollution Prediction Using Deep Learning and EEMD

Air pollution has become a hot topic in our society in recent years. Prediction of air pollution is necessary to give early warnings to the community to prevent affecting public. In this work, we utilized ensemble empirical mode decomposition and Long-short term memory to build AI model for 2 hours-ahead PM2.5 concentration prediction. Our work comprises mainly two steps: first, decomposing time-series data into intrinsic features; and one model Long-short term memory is used to learn pattern from training data. The experimental results show that our work outperform the baseline models and significantly improve the performance in term of RMSE and IOA.