Ambient air quality assessment using ensemble techniques

Air pollution is considered as an important concern all over the world. It disturbs the whole environment and produces more harmful effects to human’s healthy life. Relevant statistical reports from World Health Organization notify that air pollution play a major role in cause of diseases like asthma, lung cancer, stroke, early death and premature birth. Apart from diseases pollution also influence dangerous climate, weather conditions and may cause acid rain, global warming, ozone layer depletion, rainfall declines, etc. Therefore, it is essential to take necessary and preventive measures against air pollution. A comprehensive study is required to assess quality of ambient (outdoor) air, based on the observations of the major pollutants concentration drawn from different monitoring stations. Aiming at this problem, we proposed an ensemble based model to assess the air quality of United States from the period 2000 to 2016. In this article, we resolved the issues related to preprocessing of imbalanced dataset and improved the performance of the entire system through ensemble methods. We compared the recommended model with the existing ones. The experimental results show that the suggested model is superior to other systems and yield high accuracy, low error rate.