The electronic nose functions by comparing its sensor arrays and pattern recognition systems to the human olfactory system. It is specifically designed to detect and identify odors and flavors. The E-Nose, which emulates the human sense of smell, has attracted considerable attention and is now utilized in various fields. Air pollution poses a threat to nearly every organ in the body, with particulate matter (PM) being the primary culprit. The health effects of PM are closely linked to its particle size, which is a critical factor in its classification. In this research, the particulate matter concentrations across various urban and rural locations were evaluated using machine learning. The environmental data were collected using a developed in-house electronic nose, which incorporates a combination of sensors, including particle and gas sensors, and was validated through a series of experiments. The data was then analyzed using three selected machine learning algorithms, namely Feed Forward Neural Network (FFNN), Support Vector Machine (SVM), and k-Nearest Neighbour (kNN). The results show that the machine learning model can estimate at an accuracy of 98.59% for two-class condition of healthy or non-healthy PM reading. This study demonstrates the potential of electronic nose technology and machine learning in monitoring and evaluating air quality, and provides valuable insights into the development of more effective air pollution monitoring systems.