Particulate Matter Forecasting in Bangladesh, India, China, and USA. A Machine Learning Approach

      Аннотация: This study was an attempt to build a suitable time series model to forecast the Particulate Matter of Bangladesh, India, China, and the USA. This secondary dataset was collected from the World Bank data Indicators. There are 264 countries or regions PM2.5 air pollution, mean annual exposure (micrograms per cubic meter) recorded from 1990 to 2017. The time series plot showed that Particulate Matter had a rightly upward trend over time in Bangladesh, India, and China. But the United States of America had a rightly downward trend over time. Missing data were analyzed by the multiple imputation method. Here 'year' was the independent and Particulate Matter was the dependent variable. To fit proper Autoregressive Integrated Moving Average (ARIMA) and Generalized Autoregressive Conditional Heteroscedasticity (GARCH) family models, the stationary property of the series was confirmed. Finally, the finite mixtures of ARIMA with GARCH family models were established for forecasting purposes.

Ключевые слова: Particulate Matter, Forecasting, Air pollution, Machine Learning