• Mr. Abhishek Chadhar
• Mr. Rajneesh Pachouri
• Mr. Anurag Jain

An indicator used to gauge air quality is the index of air quality (AQI). It evaluates how air pollution may affect a person's health in the near term. The goal of the AQI is to educate the general public about the detrimental impacts of local air pollution on health. Air pollution in Indian cities has significantly increased in recent years. The air quality index can be calculated using a variety of different mathematical formulas. Numerous studies have found a link between the release of toxic waste into the air and adverse health consequences on people. Using data mining techniques is one of the more exciting ways to examine and predict AQI. The goal of this study is to identify the best effective technique for AQI prediction that will aid in climate control. One can improve the most effective method to find the best answer. As a result, in order to ensure that the problem of poor air quality is treated as effectively as possible, the work in this study involves significant research as well as the incorporation of cutting-edge techniques like SMOTE. In order to enable correct comparisons and support future researchers, it is also essential to fully illustrate and convey the metrics we utilized in our work in a way that is enlightening and instructional. Support vector regression (SVR), random forest regression (RFR), and catboost regression (CR) have all been used in the planned study to calculate the air quality index for New Delhi, Bangalore, Kolkata, and Hyderabad. In Bangalore (0.5674), Kolkata (0.1403), and Hyderabad (0.3826), Random Forest regression provides the lowest root mean square errors (RMSE), as well as higher accuracy compared to SVR and CatBoost regression for Kolkata (90.9700%) and Hyderabad (78.3672%), while CatBoost regression provides the lowest RMSE value in New Delhi (0.2792) and the highest accuracy is obtained for New Delhi (79.8622%) outcome after comparing the imbalanced datasets. With the dataset exposed to the synthetic minority oversampling technique (SMOTE) method, the outputs of Random forest regression have the lowest RMSE values in Kolkata (0.0988) and Hyderabad (0.0628), and higher accuracy is attained. The most accurate models are SVR and CatBoost regression, with the highest accuracy rates in Bangalore (90.3071%) and New Delhi (85.0847%).

AQI, Root Mean Square Error (RMSE), Random Forest Regression (RFR), Support Vector Regression (SVR), Catboost Regression (CR).

"Analysis clustering techniques with Time series and Applying PCA for ambient AQI", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.10, Issue 8, page no.c626-c636, August-2023, Available :http://www.jetir.org/papers/JETIR2308272.pdf

"Analysis clustering techniques with Time series and Applying PCA for ambient AQI", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.10, Issue 8, page no. ppc626-c636, August-2023, Available at : http://www.jetir.org/papers/JETIR2308272.pdf