Ensemble of Predictors for Forecasting the PM10 Pollution

The paper presents the novel approach to the accurate forecasting of the daily average concentration of PM10. It is based on the application of neural networks and wavelet transformation of the time series representing PM10 pollution. The main novelty of the proposed approach is the application of the ensemble of predictors, integrated using the blind source separation method or neural based integration. The numerical experiments of predicting the daily concentration of the PM10 pollution in Warsaw have shown good overall accuracy of prediction in terms of RMSE, MAE and MAPE errors.