Predicting the Air Quality Using Machine Learning Algorithms: A Comparative Study

Globally, air pollution is a severe issue that has an impact on both the environment and people’s health. Accurate air quality forecasting is essential for putting appropriate mitigation measures in place and protecting people’s wellbeing. The Air Quality Index, or AQI, is a numerical index that expresses the detrimental health implications of air pollution and the state of the air in a particular geographic region. Therefore, we use five widely recognized machine learning (ML) techniques in this study: decision tree algorithm (DT), random forest algorithm (RF), K-nearest neighbors algorithm (KNN), support vector machines (SVM), and Naive Bayes algorithm (NB) to perform the air quality forecasting. The Global Air Pollution Dataset and the AQI—Air Quality Index, which have been extracted from the Kaggle Repository which constitutes AQI values from various locations, are the two datasets on which they are implemented. Performance is assessed using four metrics: recall, F1-score, accuracy, and precision. Investigations illustrate that the random forest algorithm performs effectively in predicting air quality in both datasets.