Particle Swarm Optimisation based Machine Learning Algorithms for the Assessment of Air Quality Level

Abstract

Air pollution refers to the release of pollutants into the air that jeopardises human health and also results in serious environmental problems. Many countries are suffering from heavy air pollution. The inhalation of pollutant will lead to acute conditions like asthma, chronic lung diseases, and cancer. The Air Pollution not only harms the health of both humans and animals but also abolish the life of the plants. Hence the estimation of Air Quality is an essential. This paper employs air quality estimation using Particle Swarm Optimization. Multilayer perceptron type of back-propagation Neural Network is used to analyse the status of air pollution at various locations in India. The efficiency of the proposed model is evaluated with K-Nearest Neighbor, Random Forest and Support Vector Machine algorithms by classifying the data into six classes of pollution levels, which can be simply understood by the public as Good, Satisfactory, Moderately Polluted, Poor, Very Poor and Severe.