ANALYSIS AND PREDICTION OF AIR QUALITY INDEX IN INDIA DURING PRE AND POST COVID PANDEMIC USING MACHINE LEARNING ALGORITHMS

Abstract

Air pollution is an important environmental risk factor in propagation diseases such as lung cancer, autism, asthma and low birth weight etc. Regulation of air quality is an important task of the government in developing countries for ensuring people’s health and welfare. Air pollution differs from place to place and depends on multiple pollutant sources such as industrial emissions, heavy traffic congestions, temperature, pressure, wind, humidity and burning of fossil fuels etc. Analyzing and protecting air quality has become one of the most required activities for the government in almost all the industrial and urban areas today. In this paper, machine learning algorithms are used to analyze the concentrations of air pollutants such as SO₂, NO_x, PM_2.5, O₃ and PM₁₀. This paper analyses the air quality based on various pollutant concentrations through visualizations during pre-covid and post-covid days for effective feature extraction and decision making. A machine learning model is built using logistic regression and decision trees to predict the air quality index based on past air quality data. The experimental results show that the proposed model can be efficiently used to detect the quality of air and predictthe level of pollutants in the future.