A Novel Detection Approach of Ground Level Ozone using Machine Learning Classifiers

Pollution due to ground level ozone is one of the causes of air pollution. It is caused to human activities. It can cause several health problems. Therefore the identification of ground level ozone has become crucial. This paper has attempted to predict ozone day and non-ozone day from the dataset to make an advanced forecast so that health problems can be prevented at an early stage. Further, this research work has also attempted to detect ground level ozone using various algorithms like Support Vector Machines, K Nearest Neighbours, XGBoost, LGBM, Hist Gradient Boosting Machine and Deep Neural Networks. Finally, a thorough error analysis has been performed on these algorithms. From the result, it has been found that the Extreme Gradient Boosting (XGB) algorithm should be suitable for detection of the ground level of ozone layer as it results in 95% of accuracy.