In this article to monitor air quality, three different machine learning techniques have been used: supervised learning, unsupervised learning, and reinforcement learning. Performance of these learning techniques is measured using information about air pollution level to estimate impurity levels. Random forest regression is a supervised learning method that generated mean squared score of 0.3107; K-Means clustering, an unsupervised learning method that identifies the pollution patterns with a silhouette score of 0. 4094 and to make decisions in changing situations a reinforcement learning it is used and a total score of -2560. 36 is generated. Based on the results generated, it is found that random forest regression is good at predicting; for finding insight patterns, K-means clustering is good but reinforcement learning needs to learn more about the environments so that it can get positive rewards and can make decisions on the different environment situations. This article gives a clear summary of how different machine learning techniques can be used to monitor air quality. Based on this study, readers can choose one of these techniques to monitor air quality in an efficient way.