Abstract
Fine particulate matter (PM
2.5
) has become a prominent pollutant due to rapid economic development, urbanization, industrialization, and transport activities, which has serious adverse effects on human health and the environment. Many studies have employed traditional statistical models and remote-sensing technologies to estimate PM
2.5
concentrations. However, statistical models have shown inconsistency in PM
2.5
concentration predictions, while machine learning algorithms have excellent predictive capacity, but little research has been done on the complementary advantages of diverse approaches. The present study proposed the best subset regression model and machine learning approaches, including random tree, additive regression, reduced error pruning tree, and random subspace, to estimate the ground-level PM
2.5
concentrations over Dhaka. This study used advanced machine learning algorithms to measure the effects of meteorological factors and air pollutants (NO
X
, SO
2
, CO, and O
3
) on the dynamics of PM
2.5
in Dhaka from 2012 to 2020. Results showed that the best subset regression model was well-performed for forecasting PM
2.5
concentrations for all sites based on the integration of precipitation, relative humidity, temperature, wind speed, SO
2
, NO
X
, and O
3
. Precipitation, relative humidity, and temperature have negative correlations with PM
2.5
. The concentration levels of pollutants are much higher at the beginning and end of the year. Random subspace is the optimal model for estimating PM
2.5
because it has the least statistical error metrics compared to other models. This study suggests ensemble learning models to estimate PM
2.5
concentrations. This study will help quantify ground-level PM
2.5
concentration exposure and recommend regional government actions to prevent and regulate PM
2.5
air pollution.