Predicting the Quality of Air Using Supervised Techniques of Machine Learning

Abstract

Air contamination is the process of releasing the harmful gases into the atmosphere that are dangerous to human health and the whole planet. It is compared as one of most perilous threat that humans are never faced. It brings damage to all the animals and plants on the earth. To overcome this problem, the transport division has to analyze the air quality time to time using some machine learning techniques. Hence, predicting their quality using these techniques is became important these days. The main aim is to use classification techniques of machine learning (ML) in predicting air quality. The dataset of air quality is pre-processed with some of the techniques such as data preparing, data validation, and removal of missing values, bivariate and multivariate analysis. Now the quality of air is predicted using some supervised techniques such as Decision tree, support vector machines, Random forest, Logistic regression, K-Nearest neighbors. The various ML techniques are now compared with precision, Recall and F1 score. It is seen that decision tree performs very well than other techniques in air quality prediction. This implementation can help meteorological department in air quality prediction. In the next generation, some of the Artificial intelligence (A.I) techniques can be applied and optimized.