Prediction of Air Quality Index Using Hybrid Machine Learning Algorithm

Air pollution is an acute problem which leads to detrimental effects on human health and living conditions. Therefore, there is a need to monitor the pollution levels to inform people about the status of current air quality. This is done by an index called Air Quality Index (AQI) that maps the concentration of various pollutants into single value. To predict the AQI, a hybrid machine learning algorithm has been proposed in this paper in which the cluster classifications computed by k-means clustering algorithm are used as an input to support vector machines (SVM) algorithm. To perform the experimental work, three-year (January 2016 to January 2019) air quality data of Gurugram (Haryana) has been utilized after preprocessing it by scaling. The obtained results of the hybrid approach have been compared to the traditional SVM algorithm. Based on the empirical study, the hybrid algorithm prediction performance is better than SVM algorithm. It has been observed that the accuracy of proposed algorithm is found to be 91.25% as compared to the SVM algorithm with an accuracy of 65.93%.