AI and Machine Learning Based Classification of Air Quality Index Using COVID-19 Lockdown Period Data
Sandeep Kumar Sunori, Pradeep Juneja, Pushpa Bhakuni Negi, Sudhanshu Maurya, Pratul Raj, Deepa Nainwal
Abstract
During complete lockdown span of COVID-19 pandemic in 2020, there was a noteworthy downfall in the value of Air quality Index (AQI) due to the minimization of pollutants like SO <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf> , NO <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2</inf> , PM <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">10</inf> , PM <inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">2.5</inf> etc. A scale of 0 - 500 is used to express AQI value. An AQI of more than 300 is considered to be harmful as it has very serious consequences as far as human health is concerned. The objective of this research is to classify the air pollutant data into two categories i.e class1 (Harmful AQI) and class2 (Moderate AQI). For developing classification models, two approaches have been adopted, one is the Artificial Intelligence (AI), and other is the Machine Learning (ML). The AI and ML techniques are implemented by developing the ANN and SVM models respectively, in MATLAB. Finally, the response of both the models is evaluated, and their classification performance is compared.