The random forest (RF) algorithm is a very practical and excellent ensemble learning algorithm. In this paper, we improve the random forest algorithm and propose an algorithm called ‘post-selection boosting random forest’ (PBRF). This algorithm combines the original random forest and the Lasso method, without giving the number of decision trees for final prediction in advance, it can dynamically obtain the decision trees according to different input samples to output the prediction results. Meanwhile, we verify that the proposed algorithm can improve the performance of the model through simulation studies and real data analysis.

Assessment of the health and economic impacts of PM_2.5 pollution is of great importance for urban air pollution prevention and control. In this study, we evaluate the damage of PM_2.5 pollution using Beijing as an example. First, we use exposure–response functions to estimate the adverse health effects due to PM_2.5 pollution. Then, the corresponding labour loss and excess medical expenditure are computed as two conducting variables. Finally, different from the conventional valuation methods, this paper introduces the two conducting variables into the computable general equilibrium (CGE) model to assess the impacts on sectors and the whole economic system caused by PM_2.5 pollution. The results show that, substantial health effects of the residents in Beijing from PM_2.5 pollution occurred in 2013, including 20,043 premature deaths and about one million other related medical cases. Correspondingly, using the 2010 social accounting data, Beijing gross domestic product loss due to the health impact of PM_2.5 pollution is estimated as 1286.97 (95% CI: 488.58–1936.33) million RMB. This demonstrates that PM_2.5 pollution not only has adverse health effects, but also brings huge economic loss.

With the acceleration of urbanization, air pollution, especially PM_2.5, has seriously affected human health and reduced people's life quality. Accurate PM_2.5 prediction is significant for environmental protection authorities to take actions and develop prevention countermeasures. In this article, an adapted Kalman filter (KF) approach is presented to remove the nonlinearity and stochastic uncertainty of time series, suffered by the autoregressive integrated moving average (ARIMA) model. To further improve the accuracy of PM_2.5 forecasting, a hybrid model is proposed by introducing an autoregressive (AR) model, where the AR part is used to determine the state-space equation, whereas the KF part is used for state estimation on PM_2.5 concentration series. A modified artificial neural network (ANN), called AR-ANN is introduced to compare with the AR-KF model. According to the results, the AR-KF model outperforms the AR-ANN model and the original ARIMA model on the predication accuracy; that is, the AR-ANN obtains 10.85 and 15.45 of mean absolute error and root mean square error, respectively, whereas the ARIMA gains 30.58 and 29.39 on the corresponding metrics. It, therefore, proves that the presented AR-KF model can be adopted for air pollutant concentration prediction.