Air Quality Data Analysis And Prediction Using Modified Differential Evolution - Random Forest Algorithm

Abstract

Due to its devastating impact on human health and the environment, air pollution has emerged as a major global issue. We use the random forest technique to analyze and forecast air quality levels in this study. The purpose of this study is to use the random forest algorithm to investigate and forecast air quality in Chennai, Tamil Nadu. Three monitoring stations (Alandur Bus Depot, Manali Village, and Velachery Res. Area) provide a substantial dataset of air quality measurements used in the study. Features like meteorological data, geographical information, and temporal patterns are added to the model to enhance its forecasting powers. These characteristics supply background data that can be used to better interpret associations between air quality and other variables. Due of its stability and scalability, the random forest approach was used. The input features and air quality categories are utilized to train the model, which is then put through its paces on the testing set. The accuracy of the trained model predictions is then measured against the testing set. The findings could be used to improve Chennai air quality management. The proposed model using Modified Differential Evolution based Random Forest (MDE-RF) can help policymakers, environmental agencies, and public health groups make better decisions and put into action more effective actions to reduce the effects of air pollution. In addition, it provides new opportunities for studying the dynamics of air pollution in metropolitan areas and developing cutting-edge machine learning algorithms for air quality prediction.