Abstract
Air pollution poses significant risks to human health, the environment, and the economy. Therefore, striving for cleaner air through efficient air quality monitoring is imperative for fostering a healthier and more sustainable future. Predicting air quality is essential to enhance the quality of life, maintain environmental sustainability, and reduce the economic burden associated with poor air quality issues. The artificial neural network (ANN) is widely recognized as a predominant computational tool in air quality studies due to its capabilities in predicting gaseous and particulate pollutant concentrations, as well as forecasting the air pollutant index (API). This study aimed to investigate the predictive performance of ANN in determining the API by utilizing identified potential sources of air pollutants. Five prediction models were created, namely ANN-PC2018, ANN-PC2019, ANN-PC2020, ANN-PC2021, and ANN-PC2022. Principal component analysis (PCA) was conducted to identify the most significant sources of air pollution, and the results were employed to predict the API using ANN. The ANN-PC2019 model exhibited the highest performance with an R2 value of 0.8612 and RMSE of 7.7467, utilizing four major pollutants as input variables. These findings suggest that forecasting air quality using fewer parameters yields reliable outcomes.