The Predictive analysis aims to generate a prediction about futuristic trends based on present events. This process includes different steps of data pre-processing, feature extraction and regression. The dataset which is used for the prediction analysis has various attributes and various instances which are used for the prediction. The datasets are downloaded from the authentic source. When the dataset has various missing and redundant values, it is pre- processed to remove those values. In the second phase, the relationship between various attributes and target is established to check which attribute has maximum impact on the target set. For air quality prediction, Linear, LASSO, Elastic net, and Random Forest tree regression are used. Where the number of attributes is high, these regression models are unable to correctly forecast air quality. For more accurate air quality estimation, voting regression, a combination of linear, elastic net, and decision tree regression, is used. In comparison to other existing regression models, the proposed model produces improved outcomes in terms of RMSE and R2, which are 1.78 and 0.95, respectively.