A Machine Learning Framework for Forecasting Air Quality Index

Air pollution involves complex interactions between meteorological factors, emissions, geographical features, and human activities. Traditional statistical models often struggle to capture these non-linear and multi-dimensional relationships, while Machine Learning (ML) and Deep Learning (DL) models are superior at this. Accurate predictions and insights from ML and DL models enable policymakers to implement timely interventions, such as issuing health advisories, optimizing traffic, and controlling industrial emissions. This study explores the application of ML and DL models to predict Air Quality Index (AQI) using a dataset collected from one of air polluted cities in the world - Tashkent, the capital of Uzbekistan. The problem is analyzed a classification task using the K-Nearest Neighbors (KNN), Random Forest (RF), Decision Tree (DT), Support Vector Machine (SVM), and Artificial Neural Network (ANN) with different optimizers such as Adam, AdaGrad, RMSprop, SGD with momentum. Performance metrics, including Accuracy, Precision, Recall, and F1-Score are utilized to evaluate model performance. The results exhibit the strengths and limitations of different models and offer insights into the most suitable approaches for AQI prediction.