ABSTRACT
During every winter, causalities with disorders via air contamination are rapidly increasing. WHO's 2018 report states that around 7 million individuals die every 12-month period as a result of diseases attributable to air pollution. Due to this one grows extraordinarily crucial towards precisely monitoring and anticipating the grades appropriate contaminants manifested surrounded by this air. The proposed work is to predict AQI accurately with utilizing the data set on different ML model with proper pre-processing technique for finding nature concerning the air rests for the most part signified by its AQI (air quality index) value. In this chapter, it is tried to gauge the air condition in the bounds of a definite zone by utilizing machine learning strategies like support vector regression (SVR), decision tree regression (DTR), multiple linear regression (MLR) and random forest regression (RFR). It has been concluded that which machine learning method are predicting the concerned air nature accurately and all the machine learning models are studied and compared with the use of assorted error metrics, for instance, coefficient of determination (R2), root mean square error (RMSE), mean absolute error (MAE) and root mean square logarithmic error (RMSLE). The basic outcomes from the models are well analysed and shown that results given by SVR were poor. MLR and DTR were both done acceptably well. RFR played out the finest among all regression examples.
