A Machine Learning Approach for Predicting Air Quality Index in Smart Cities

One essential natural resource, the air quality index has been deteriorated by economic activities. Many studies have been conducted about the forecasting of periods of unsafe atmospheric conditions; however, the majority of these studies are limited in their capacity to include seasonal and other factors due to insufficient longitudinal data. India noise pollution has compiled a 6-year dataset that has been used to build several prediction models. There are encouraging outcomes when AQI levels are predicted using machine learning methodologies such as support vector machines (SVMs) artificial neural networks (ANNs), and Naive Bayes (NB). However, in several tests involving datasets from three different locations, NB, ANN, and SVM performed well to maximize prediction performance. When it comes to mean absolute error (MAE), R-squared (R2), and Root-mean-square deviation (RMSE), the combined performance of stacks is unsurpassed. In the related work section, summarize the limitations of existing models, such as inadequate data preprocessing, low spatialtemporal resolution, limited feature selection, or overfitting in complex models. Highlight gaps in scalability, real-time adaptability, and generalization across regions, emphasizing the need for robust, efficient models tailored to air quality prediction.