Predicting Air Quality Index in Real Time and Classifying Its Health Effects: Advancements in Machine Learning

Abstract

Accurate and early forecast of air quality is therefore very important since air pollution seriously compromises public health and environmental sustainability. This work includes a thorough investigation on real-time Air Quality Index (AQI) prediction utilizing advanced machine learning approaches and classification of its health impacts. To build strong predictive models, the proposed method combines several environmental information including meteorological factors and pollution concentrations. Urban and rural environments are guaranteed the real-time framework's applicability since it is maximized for scalability and reactivity. The study also assesses the performance of the model by means of large-scale experiments, therefore contrasting interpretability, latency, and accuracy. The results show how well machine learning might improve air quality monitoring systems and support proactive actions to lower health hazards. This study emphasizes how important artificial intelligence is in tackling public health and environmental concerns of great relevance. The quality of the air gets daily declining in recent years. To control and stop regional air pollution, exact forecast of AQI concentration is really useful. It is used to quantify how human health suffers under air pollution. One can find the air quality index in numerous methods. To find the AQI in this work we apply PM2.5, PM10, NO, NO2, NOX, NH3, CO, SO2, O3, benzene, Toluene characteristics. Among the most fascinating approaches to forecast and evaluate AQI are machine learning methods. This work applies XGBoost Regressor, Catboost Regressor, Random Forest Regressor algorithms. One can hone the most effective technique to identify the ideal answer. Thus, the work of this article consists on thorough investigation and application of new technologies such SMOTE to guarantee the best feasible solution to the issue of air quality.