An Experimental Comparison on Machine Learning Ensemble Stacking‐Based Air Quality Prediction System

Air is most important factor for human life. Other than human life is wildlife and plants are depending on air for their survival. Air is polluted by the human behavior, industrialization, and urbanization. Prevention of this air pollution has become a necessary action in many cities. Air is polluted at intolerable levels by industries and heavy vehicular traffic in cities which affects human health conditions to a great extent. Forecasting and controlling the air pollution is the need of the hour to care for human beings from health hazards. The most important objective of this paper is to proposed new method to predict air pollution using data collected on monthly basis and give recommendations to prevent and control air pollution. Ambient air monitoring is the regular, continuing measurement of pollutant levels by measuring the amount and types of certain pollutants in the outside air. Investigative and protecting air quality in this earth has become one of the fundamental activities for every human in many industrial and urban areas at the present time. Based on the serious health concerns and the atmospheric pollution has become a main source of premature mortality among general public by causing millions of deaths for every year based on (WHO, 2014). Air Quality prediction used to alert the people about the air quality alarming conditions, and it's health effects and also support Environmentalists and Government to frame air quality standards and regulations based on issues of harmful and pathogenic air exposure and health‐related issues for human welfare. Air Quality Index (AQI) is a measure of pollution level in the air. Predicting air pollution with AQI is one the major challenging area of Research nowadays. Machine Learning (ML) methods are used to predict the AQI. Machine Learning methods is a scientific approach to solve certain tasks and predict the value using techniques and Algorithms such as Supervised Learning (SL), Semi Supervised Learning (SSL), and Unsupervised Learning (USL). Machine Learning algorithms provide various methods to forecasting the air pollution levels. Ensemble method of Machine Learning algorithms is applied to predict the air quality and analyze these results to conclude with the comparison of other regression algorithms.