Covid-19 Salgın Sürecinde Hava Kalitesi Tahmini: Zonguldak Örneği

Abstract

Air pollutants are known to cause severe effects on human health, from simple effects to premature death. It causes many ailments, especially respiratory problems, lung diseases, and pneumonia. As of December 30, 2019, the covid-19 outbreak affecting the world is a respiratory disease and is transmitted by air. In the current epidemic process, it is essential to predict air quality and take measures to affect the rate of spread of air-borne diseases such as Covid-19. The study predicts air quality through machine learning methods, considering the various concentrations of pollutants measured before the covid-19 outbreak and during the covid-19 outbreak. The data set used in the study consists of the pollutant concentrations of Zonguldak province, which has high air pollution and developed industry. The data was obtained from the Ministry of Environment and Urban Planning (MoLS) weather monitoring stations website. Five different machine learning methods with high predictive success were used. As a result of the study, the best hit was achieved in the decision tree algorithm with Rmse values of 0.016 (2019 dataset) and 0.021 (2020 dataset). The Naive Bayesian algorithm has the lowest success in the study. Experimental results suggest that the proposed model could be used efficiently to detect air quality. (English) [ABSTRACT FROM AUTHOR]