Air pollution analysis using enhanced K-Means clustering algorithm for real time sensor data

Air pollution affects body organs and human systems in addition to the environment. Smart air pollution monitoring consists of wireless sensor nodes, server and a database to store the monitored data. Huge amounts of data are generated by gas sensors in air pollution monitoring system. Traditional methods are too complex to process and analyze the voluminous data. The heterogeneous data are converted into meaningful information by using data mining approaches for decision making. The K-Means algorithm is one of the frequently used clustering method in data mining for clustering massive data sets. In this paper, enhanced K-Means clustering algorithm is proposed to analyze the air pollution data. The correlation coefficient is calculated from the real time monitored pollutant datasets. The Air Quality Index (AQI) value is calculated from the correlation co-efficient to determine the air pollution level in a particular place. The proposed enhanced K-Means clustering algorithm is compared with Possibilistic Fuzzy C-Means (PFCM) clustering algorithm in terms of accuracy and execution time. Experimental results show that the proposed enhanced K-Means clustering algorithm gives AQI value in higher accuracy with less execution time for when compared to existing techniques.