Air pollution has become a global environmental problem, because it has a great adverse\nimpact on human health and the climate. One way to explore this problem is to monitor and predict\nair quality index in an economical way. Accurate monitoring and prediction of air quality index (AQI),\ne.g., PM2.5 concentration, is a challenging task. In order to accurately predict the PM2.5 time series,\nwe propose a supplementary leaky integrator echo state network (SLI-ESN) in this paper. It adds the\nhistorical state term of the historical moment to the calculation of leaky integrator reservoir, which\nimproves the influence of historical evolution state on the current state. Considering the redundancy\nand correlation between multivariable time series, minimum redundancy maximum relevance\n(mRMR) feature selection method is introduced to reduce redundant and irrelevant information,\nand increase computation speed. A variety of evaluation indicators are used to assess the overall\nperformance of the proposed method. The effectiveness of the proposed model is verified by the\nexperiment of Beijing PM2.5 time series prediction. The comparison of learning time also shows the\nefficiency of the algorithm.
Loading....