Organic aerosol (OA) accounts for a significant fraction of atmospheric particulate matter. The OA concentration in the atmosphere is of high variability and depends on factors such as emission, the atmospheric oxidation process, meteorology, and transport. Due to the complex interactions among the numerous factors, accurate estimation of the effects of target variables on OA concentration is often challenging. Herein, a random forest machine learning algorithm successfully predicted the concentrations of primary and oxygenated organic aerosol (POA and OOA) at urban and rural sites in Hong Kong. The random forest model explained more than 80% of the observed traffic-POA, cooking-POA, and OOA. In contrast, a multiple linear regression model only explained 30-50% of these OA concentrations. In the random forest model training process, NOwas also the most important variable for traffic-POA and cooking-POA. For OOA, multiple parameters were equally crucial in the model prediction, including NO, O, and relative humidity (RH). The dependence of OA concentrations on atmospheric conditions (e.g., various NOand Oconcentrations and meteorological conditions) was calculated via the partial dependence algorithm. The results suggested that the dependence of OA concentrations on atmospheric conditions was nonlinear and depended on different condition regimes. The partial dependence algorithm provides insights into the POA source and OOA formation mechanisms under a complex environment.