Air is one of the most fundamental elements required by all living beings to survive on the planet. Our planet Earth is surrounded by a thin layer of air that extends up to several kilometers above the surface of the earth and that forms our atmosphere. Air is a mixture of various components like nitrogen (N2) 78.08%, oxygen (O2) 21%, carbon dioxide (CO2) 0.04%. Rest of the components like water vapor, argon, dust, and smoke, are present in very small amounts. These are the components that play a major role in determining the quality of the air that we have around us. When the proportions of these components get disturbed, this leads to air pollution. Air pollution has become a major environmental problem affecting the health of people worldwide. In recent years, due to the availability of air pollution data, time series mining techniques has emerged as a promising approach predicting air pollution levels by identifying underlying trends, patterns and also forecasting the various components present in air.  Many researchers are continuously working to analyze the air quality. Several methods, like analytical, statistical, and ensemble methods, have been adopted for getting accurate results for the analysis and forecasting of the air quality. In this paper, we have studied the works of several scholars, the models proposed by them, an analysis made by them, and the predictions they have made for forecasting air quality. This paper presents a concise literature summary of the works and focuses on their research gaps which may be helpful for further studies and analysis of air quality.

___________________________________________________________________________________

Keywords: Air Quality, Boosting Algorithms, Forecasting Models, Predictors, Time Series data.