Prediction of Particulate Matter (PM2.5) for Industrial Area Based on Naive Bayes Classifier

In India, the industry is gaining more attention from the public, and thus the correlation between factories and nearby pollution sources has done a subject of research. Industrialization and urbanization are increasingly expanding to make India a developed country. The outcome of modern automation contains more droplets, gas molecules, and solid particles that can lead to medical issues when inhaled. This research focuses on a comparative analysis of two machine learning algorithms to predict PM2.5 concentration monitored in an industrial area. The random forest and Naïve Bayes classifiers have been compared to predict the class of PM2.5 concentration monitored in the industrial area of Haridwar City (SIDCUL). Research shows that the Naïve Bayes classifier is best with an accuracy of 97.37% to predict the class of PM2.5 pollutants. The study shows that the concentration of PM2.5 is low in winter compared with other seasons. This prediction model of PM2.5 concentration will be proven helpful for scientists and researchers, and information about PM2.5 concentration for this industrial area will also become important for the government.