Air quality index prediction for Gorakhpur city using k-nearest neighbors: Model evaluation and analysis

Emissions have increased and city air quality requirements have decreased due to rapid urbanization. Living in a city is negatively impacted by the increasing levels of toxins in the air. In these cities, the ambient air quality is measured and reported by the CAAQMS based there. This work delves deeper into applying models based on machine learning for AQI prediction. Gorakhpur City, Uttar Pradesh, India's CPCB provided the source data for this study. NO2, SO2, and particle matter (PM10 and PM2.5) were the primary AQI pollutant measurements. This study examines the effects of Gorakhpur City's AQI on health using the K-Nearest Neighbors (KNN) method. The model finds patterns and relationships between emissions and respiratory ailments by combining temporal and spatial data on traffic density, pollutant concentrations, and climatic conditions. Recognized for its simplicity and efficacy, the KNN model forecasts possible health hazards and classes areas with high pollution levels. The results provide useful information about the KNN model that it can develop a robust model as it generates lower evaluation metrics and higher coefficient of determination, so they may put specific pollution reduction and public health protection policies into action. The model shows higher accuracy with an R2 value of 0.985. which indicates the model's capability to recognize the larger variance in the dataset. With its machine learning tool, we can develop a robust model to forecast AQI even with a small dataset.