PREDICTING AIR QUALITY USING HISTORICAL IOT SENSOR DATA AND ENSEMBLE LEARNING APPROACHES

International Journal of Computer Science (IJCS) Published by SK Research Group of Companies (SKRGC)

Abstract

With the rising levels of air pollution in urban areas, there is a critical need for accurate forecasting systems that can anticipate poor air quality conditions before they pose health risks. This research introduces a predictive model that leverages historical data from commercial IoT-based gas sensors to estimate the Air Quality Index (AQI) using ensemble machine learning techniques. The study primarily focuses on pollutants such as CO, SO?, NO?, and NH?, and includes sensor calibration to enhance the reliability of the collected data. A Random Forest Regressor was implemented as the core ensemble learning algorithm, selected for its ability to handle complex data patterns and reduce prediction error. The model achieved a high accuracy of 91.2%, with a precision of 90.7% and recall of 91.8%, indicating strong performance in classifying AQI categories. The RMSE value of 4.72 suggests minimal deviation between actual and predicted AQI values. Visual evaluations through confusion matrix and heatmap confirmed that most predictions were correct, with a few misclassifications occurring near category boundaries—an expected outcome due to the closeness of threshold values. This study demonstrates the potential of integrating calibrated IoT sensor data with ensemble learning models to build reliable AQI forecasting systems, supporting smarter environmental monitoring and urban planning initiatives.

References

1. A. Rajendran, R. Srinivasan, and S. Kumar, "IoT-based air quality monitoring system," IEEE Sensors Journal, vol. 16, no. 8, pp. 2345-2351, 2016.

2. J. Gao, L. Li, and Y. Zhang, "A real-time air quality monitoring system based on IoT sensors for urban environments," Environmental Science and Pollution Research, vol. 26, no. 12, pp. 11957-11966, 2019.

3. M. Jin, X. Zhang, and H. Li, "Multi-sensor network for air quality monitoring: Applications and developments," Sensors and Actuators B: Chemical, vol. 240, pp. 1066-1075, 2017.

4. X. Zhou, Z. Wu, and F. Yu, "A hybrid machine learning approach for air quality prediction," Journal of Environmental Management, vol. 218, pp. 345-356, 2018.

5. S. Patel, A. Agarwal, and M. Sharma, "A machine learning approach for AQI prediction based on meteorological data," Environmental Monitoring and Assessment, vol. 192, no. 10, pp. 1-10, 2020.

6. W. Zhang and T. Wang, "A multi-sensor data fusion approach for AQI prediction," Journal of Environmental Informatics, vol. 27, no. 3, pp. 225-235, 2016.

7. J. Liu, F. Zhao, and S. Sun, "Ensemble learning for air quality prediction using multiple machine learning techniques," Science of the Total Environment, vol. 650, pp. 1043-1051, 2019.

8. X. Chen, H. Xu, and Y. Li, "AQI forecasting using gradient boosting machine," Environmental Science and Technology, vol. 54, no. 12, pp. 8236-8244, 2020.

9. M. Khan, M. Rehman, and S. Choi, "Ensemble learning approach for air quality prediction," Springer Environmental Science and Pollution Research, vol. 25, no. 3, pp. 2397-2405, 2018.

10. I. Hussain, M. Qureshi, and S. Farhan, "Deep ensemble learning for AQI prediction," IEEE Transactions on Neural Networks and Learning Systems, vol. 31, no. 6, pp. 2021-2031, 2020.

11. M. Hossain, M. Alam, and S. Rahman, "Air quality prediction using feature extraction and machine learning," Sensors, vol. 19, no. 11, pp. 1-14, 2019.

12. A. Singh, A. Shukla, and S. Pandey, "Cloud-based real-time air quality prediction using IoT sensors," Environmental Pollution, vol. 229, pp. 456-463, 2017.

13. S. Amin, S. Hu, and M. Lin, "Cloud-based predictive model for AQI using IoT data," Computers, Environment and Urban Systems, vol. 68, pp. 35-42, 2018.

14. A. Siddique, S. Rashid, and S. Khan, "Challenges in IoT-based air quality monitoring systems," Environmental Monitoring and Assessment, vol. 193, no. 1, pp. 1-15, 2021.

15. M. Xie, T. Zhang, and Y. Zhang, "Real-time air quality prediction with IoT sensors: Issues and solutions," Journal of Environmental Science and Technology, vol. 54, no. 9, pp. 5734-5742, 2020.

16. S. Rashid, M. Shah, and A. Khan, "IoT-based real-time air quality monitoring system," Environmental Science and Pollution Research, vol. 25, no. 4, pp. 3547-3556, 2018.

17. A. Kumar and R. Gupta, "Prediction of AQI using machine learning algorithms," Journal of Environmental Informatics, vol. 23, no. 3, pp. 197-206, 2019.

18. A. Alam, M. Rahman, and S. A. Ali, "Ensemble learning methods for AQI prediction," Environmental Monitoring and Assessment, vol. 192, no. 12, pp. 1-10, 2020.

19. Y. Zhao, J. Wang, and J. Liu, "AQI prediction using deep learning models," Journal of Environmental Management, vol. 218, pp. 345-356, 2017.

20. S. Chakraborty, A. Saha, and S. Dey, "Data fusion for air quality prediction using multiple sources," Environmental Science and Technology, vol. 54, no. 15, pp. 9690-9699, 2020.

21. UCI Machine Learning Repository, “Air Quality Dataset,” Kaggle, 2020.

Keywords

Air Quality Index (AQI), IoT Sensors, Ensemble Learning, Random Forest Regressor, Sensor Calibration, Environmental Monitoring.