The automated classification of ambient air pollutants is an important task in air pollution hazards assessment and life quality research. Faced with various classification algorithms, environmental scientists should select the most appropriate method according to their requirements and data availability. This study describes several types of Decision Tree algorithms for finding the inter-correlation between dominant air pollution index (API) for PM10 percentile values and four other air pollutants such as Sulphur Dioxide (SO2), Ozone (O3), Nitrogen Dioxide (NO2) and Carbon monoxide (CO), in addition to two other meteorological parameters: ambient temperature and humidity, using 22 months records of active air monitoring station in Penang island (northern Malaysia). Classification analysis for the PM10 API was then performed using non-linear Decision Trees within the R programming environment including: Boosted C5.0, Random Forest, PART, and Naive Bayes tree (NBtree). This is in addition to rpart and tree algorithms, which were used to plot the classification trees. The classification performance of the methods is presented and the best classifier in terms of accuracy and processing time was recommended. In R statistical environment, the process of classification by decision tree methods and the classification rules were easy to obtain, while geographic information systems (GIS) software' was used for mapping the study area. Furthermore, the results are clear and easy to understand for environmental and geospatial scientists and relevant agencies, which will facilitate the mitigation of air pollution related disasters in the affected communities. © 2020 IOP Publishing Ltd. All rights reserved.