Abstract
Air pollution has thus exceeded the anticipated safety limit due to the growing usage of automobiles, the manufacturing sector, and the production of pollutants from other human activities. It is considered one of the major environmental risks to humankind on Earth. Nowadays, monitoring and forecasting air quality is significant, particularly in high-level pollution countries. In contrast to traditional methodologies, predictive technologies based on machine learning approaches be the most effective instruments for analyzing such contemporary threats. Therefore, this paper presents multiclass classification using two feature selection techniques namely Sequential Forward Selection (SFS) and Filter with different Machine Learning and Ensemble techniques to predict the air quality. Therefore, intensive research is conducted in novel techniques such as Wrapper and Filter feature selection methods to make sure that the most relevant features are on datasets for the solution of the air quality problem. The results of the considered framework reveal that the Wrapper feature selection technique provides superior performance compared to various Filter feature selection with different ML methods, including AdaBoost Classifier, Extra Tree Classifier, KNN, RF, GB, and Bagging Classifier for efficiently determining the Air Quality Index (AQI). Its important goal is to visualize the air quality datasets to understand and see the hidden sight in datasets. These models' performances are assessed and compared using predetermined performance metrics. The AdaBoost Classifier model with Filter selection has the lowest accuracy, while the Random Forest Classifier model with Wrapper feature selection achieves the highest accuracy with 78.4% and 99.99% respectively. Based on the raw data set, it was noted that the F1-score, Recall, and Precision values of the Random Forest model Wrapper Feature selection achieve 99.96%, 99.97%, and 99.98% respectively. Therefore, the experimental results undoubtedly show the supremacy of the proposed approach, providing a practical, reliable, and robust tool to effectively determine the Air Quality Index (AQI)