Enhanced Stacking Ensemble Model in Predictive Analytics of Environmental Sensor Data

Abstract— Predictions based on existing data are an effective way to protect human lives. It plays a vital role in taking precautionary measures and minimize the degree of damage. In India, air pollution is a major issue causing several health problems like respiratory difficulties, lung cancer and even cardiopulmonary deaths. Air is contaminated by various pollutants, among which Particulate Matter(PM2.5) is known to be the toxic particles smaller than 2.5 micrometers in diameter. This paper focuses on prediction of PM2.5 concentration level in the air using an enhanced stacking ensemble machine learning model. The experimental outcome indicates our proposed model performs better comparative to other ensemble models.