AQI-Cast: An Approach for Air Quality Index Prediction Using Hybrid Deep Learning Models

Air pollution today is one of the most critical global challenges in recent decades, primarily driven by industrialization, urbanization, and an ever-growing population. The deteriorating air quality, marked by the increased concentration of pollutants such as Carbon Monoxide (CO), Nitrogen Dioxide ($\text{NO}_{2}$), Sulfur Dioxide ($\text{SO}_{2}$), and particulate matter ($\text{PM}_{2.5}$ and $\text{PM}_{10}$), poses dangerous threats to human health, particularly the cardiovascular and respiratory systems. Accurate Air Quality Index (AQI) prediction has therefore become imperative for devising timely mitigation strategies, safeguarding public health, and enabling informed urban planning and policymaking. Our paper aims to accurately predict air quality index from a public dataset ensuring high accuracy scores. Our study involves using deep learning networks and hybrid models to perform the given task.