Air pollution poses a serious risk to environmental sustainability and public health, especially in urban agglomerations. Accurate air quality prediction and early anomaly detection are crucial for effective environmental management and preventive medicine. This paper proposes a deep learning-based method that integrates Convolutional Neural Networks (CNNs) for air quality level prediction and DBSCAN clustering for identifying anomalous pollution patterns. A dataset spanning five years (2018–2022) consisting of 2,797 samples and 13 attributes was utilized for the training and evaluation of the models. The dataset was divided into 70% for training, 15% for validation, and 15% for testing. The CNN model was trained for 50 epochs using the Adam optimizer with has a batch size of 32, employing Mean Squared Error (MSE) as the loss function. The model's performance was assessed using standard metrics, including Accuracy, Precision, Recall, and F1-score findings demonstrate a consistent and dependable performance, achieving an overall accuracy of 76.0%, precision of 77.5%, recall of 76.2%, and an F1-score of 76.3%. The confusion matrix also indicated areas of strength and weakness, particularly the impact of false negatives on public safety. With a view of improving anomaly sensitivity, this currently proposed technique in this study, has the potential to be used in real-time monitoring of air quality and policy support.