Research on Prediction of Air Quality CO Concentration Based on Python Machine Learning

With the accelerating pace of urbanization, the issue of air pollution has become increasingly severe. Notably, carbon monoxide (CO), as a prevalent harmful gas, poses potential threats to both human health and the environment. Therefore, accurate prediction of CO concentration and analysis of its influencing factors are of significant importance for urban environmental management and public health protection. This study utilizes air quality monitoring data from the UCI open database, selecting multidimensional features including gas sensor outputs and meteorological conditions, and employs a Random Forest regression model to predict CO concentrations. By comparing actual values with predicted values, the model’s performance was evaluated using Mean Absolute Error (MAE) and the Coefficient of Determination (R²). The results indicate that the proposed method can, to some extent, accurately reflect the variation trends of CO concentrations. Furthermore, through feature importance analysis, it was found that features such as benzene concentration (C6H6 (GT)), nitrogen oxides (Nox (GT)), nitrogen dioxide sensor readings (PT08.S4 (NO₂)), and carbon monoxide sensor readings (PT08.S1 (CO)) exhibit high contributions in predicting CO concentrations. This research provides a valuable reference for air pollution prediction and intelligent environmental governance.