Evaluation of Machine Learning Algorithms for Air Quality Index (AQI) Prediction

The Air Quality Index (AQI) has been deteriorated due to the growth of industry and automobiles in many regions of India. Artificial intelligence and machine learning have greatly benefited the ability to predict air quality. This paper aims to know the status of air pollutants (PM10, PM2.5, SO2, and NO2) monitored in different cities of Uttarakhand State (India) and the Air Quality Index (AQI) using the Python language (Jupyter Notebook). The air quality index dataset has used six machine-learning algorithms (Logistic Regression, Naive Bayes, Random Forest, Support Vector Machine, K-Nearest Neighbors, and Decision Tree). These machine-learning algorithms have been evaluated based on precision, recall, accuracy, etc. The result shows that Random Forest and Decision Tree algorithms outperformed each other and achieved the highest accuracy, i.e., 99.0%. Further, the air quality index (AQI) values have also been predicted and compared to actual values using the random forest algorithm.