A Hybrid Air Quality Prediction Model Based on Empirical Mode Decomposition

Air pollution is a severe environmental problem in urban areas. Accurate air quality prediction can help governments and individuals make proper decisions to cope with potential pollution. As classic time series forecasting model, the AutoRegressive Integrated Moving Average (ARIMA) has been widely adopted prediction. However, because of volatility lack additional context information, i.e., spatial relationships among monitor stations, traditional ARIMA models suffer from unstable performance. Though some deep networks achieve higher accuracy, mass training data, heavy computing, cost are required. In this paper, we propose hybrid model simultaneously predict seven indicators multiple monitoring stations. The proposed consists three components: (1) an extended matrix several adjacent stations; (2) Empirical Mode Decomposition (EMD) decompose data into smooth sub-series; (3) truncated Singular Value (SVD) compress denoise expanded matrix. Experimental results on public dataset show that our outperforms state-of-art both accuracy cost.