CG-CPO Optimized Voting Ensemble Framework for Accurate Air Quality Index Prediction using Environmental and Pollutant Data
Abstract
Accurate forecasting of air pollution is important for monitoring the environment and protecting human health. The Air Quality Index (AQI) is commonly used to measure pollution levels, but forecasting it proves to be challenging because of the complex interactions that occur between contaminants, weather conditions, and changes that happen over time. This study utilizes machine learning (ML) techniques to predict the Air Quality Index (AQI) based on pollution and environmental data. The method uses feature analysis, data pre-processing, and an ensemble voting regression model. Chaotic-Guided Crested Porcupine Optimizer (CG-CPO) is used to find the best values for the parameters. The model achieves low error values with a Mean Absolute Error of 5.72 and a Root Mean Square Error of 8.71, alongside a high R2 score of 0.981, indicating strong results. To facilitate understanding of the data, AQI values are categorized into standard air quality classifications. This approach offers a useful and reliable method for monitoring air quality and making decisions based on that information.