Potential of ARIMA-ANN, ARIMA-SVM, DT and CatBoost for Atmospheric PM2.5 Forecasting in Bangladesh

Atmospheric particulate matter (PM) has major threats to global health, especially in urban regions around the world. Dhaka, Narayanganj and Gazipur of Bangladesh are positioned as top ranking polluted metropolitan cities This study assessed performance application hybrid models, that is, Autoregressive Integrated Moving Average (ARIMA)-Artificial Neural Network (ANN), ARIMA-Support Vector Machine (SVM) Principle Component Regression (PCR) along with Decision Tree (DT) CatBoost deep learning model predict ambient PM2.5 concentrations. The data from January 2013 May 2019 2342 observations were utilized this study. Eighty percent was used training rest dataset employed testing. models evaluated by R2, RMSE MAE value. Among performed best for predicting all stations. values during test period 12.39 µg m?3, 13.06 m?3 12.97 Gazipur, respectively. Nonetheless, ARIMA-ANN DT methods also provided acceptable results. suggests adopting atmospheric Bangladesh.