Air Quality Prediction and Prevention Recommendations based on Regression Algorithms

In the context of promoting carbon neutrality policies, air quality prediction has become a key link in environmental governance. This study used Python language to clean and perform feature engineering on the air quality monitoring data of Kaggle platform, and established a prediction model using multiple regression algorithms. The study found that the ExtraTreesRegressor model performed the best, with a MAE of 2.829, MSE of 11.243, R² of 0.942, and an accuracy of about 88%. CO, PM2.5, and NO2 were identified as the main factors affecting AQI. The research results can provide reference for air quality prediction and environmental governance.