Machine learning-based forecasting of air quality index under long-term environmental patterns: A comparative approach with XGBoost, LightGBM, and SVM

Sevtap Tırınk ¹

⁰Department of Medical Services and Techniques, Environmental Health Program, Iğdır University, Iğdır, Türkiye.

Plos One +

|

October 8, 2025

View abstract on PubMed

Summary

Machine learning models accurately predict daily Air Quality Index (AQI) using meteorological and pollutant data. eXtreme Gradient Boosting (XGBoost) demonstrated superior performance, offering insights for air pollution management and public health protection.

Area Of Science

Environmental Science
Data Science
Public Health

Background

Air pollution poses a significant global threat to environmental sustainability and public health.
Accurate air quality monitoring and prediction are essential for policy development and public safety measures.

Purpose Of The Study

To assess the long-term daily Air Quality Index (AQI) prediction capabilities of machine learning models.
To evaluate the effectiveness of meteorological and air pollutant data in predicting AQI in eastern Türkiye.

Main Methods

Utilized daily data from 2016-2024, including PM₁₀, SO₂, NO₂, O₃, temperature, precipitation, humidity, wind direction, and wind speed.
Compared three machine learning models: eXtreme Gradient Boosting (XGBoost), Light Gradient Boosting Machine (LightGBM), and Support Vector Machine (SVM).
Performance was evaluated using R², RMSE, and MAE metrics.

Main Results

XGBoost achieved the highest prediction accuracy with R² = 0.999, RMSE = 0.234, and MAE = 0.158.
Ensemble machine learning methods, especially XGBoost, effectively modeled AQI variations based on environmental factors.
The study highlights the potential of machine learning for robust air quality forecasting.

Conclusions

Machine learning models, particularly XGBoost, offer a highly accurate approach to daily AQI prediction.
Findings support the development of advanced air quality forecasting systems and regional pollution management strategies.
The study provides practical insights for enhancing public health protection through early warning systems and environmental policy.