Article Details

Abstract

This paper studies and analyzes prediction models based on different Machine Learning methods aiming to improve the accuracy of urban air quality prediction. Taking the air quality of Guangzhou in 2018 as an example, firstly, the correlation between urban air quality and its characteristics is analyzed through time series and correlation research. Then, based on Random Forest regression, Decision Tree regression and Gradient Boosting Tree algorithm respectively, AQI prediction models are constructed to determine the better model. Finally, the model is optimized by grid search for parameter tuning. The results show that the best prediction model is based on Random Forest regression, with a RMSE of 8.93 and a goodness of fit of 0.88. It has strong prediction accuracy and can effectively predict the urban air quality index.

Keywords

urban air quality; Machine Learning; grid search; prediction model

References

1. [1] PIETRARU R N，NICOLAE M，MOCANU S，et al. Easy-to-Use MOX-Based VOC Sensors for Efficient Indoor Air Quality Monitoring [J/OL].Sensors，2024，24（8）：2501（2024-04-13）[2024-09-15].https://doi.org/10.3390/s24082501.
2. [2] MÉNDEZ M，MERAYO M G，MANUEL N. Machine Learning Algorithms to Forecast Air Quality: A Survey [J].Artificial Intelligence Review，2023，56（9）：31-36.
3. [3] GUTTIKUNDA S K. Designating Airsheds in India for Urban and Regional Air Quality Management [J].Air，2024，2（3）：247-257.
4. [4] Wuhan Academy of Social Sciences Research Group Comparative Study on the Development of Wuhan and Guangzhou under the New Development Pattern [J]. Wuhan Social Sciences, 2023 (2): 37-46
5. [5] Guo Sandang, Li Qian, Jing Yaqian Comprehensive evaluation of urban air quality based on improved TOPSIS method [J]. Henan Science, 2021, 39 (11): 1842-1849
6. [6] Zhu Feihong, Chen Wensheng, Hou Shuai, etc Research on Air Quality Impact Factor Analysis Based on RFA [J]. Computer Simulation, 2020, 37 (10): 221-225
7. [7] Zhang Qin, Guo Jinli Research on Machine Learning based Air Quality Prediction Method in Shanghai [J]. Software Guide, 2022, 21 (8): 33-38
8. [8] Xiao Liang, Zeng Yun RUL prediction of turbofan engine based on decision tree feature extraction and RSM-LightGBM [J]. Agricultural Equipment and Vehicle Engineering, 2024, 62 (4): 132-136
9. [9] Liu Guanshen, Zhang Yan, Cao Xin, etc Research and Application of Random Forest Method in PM2.5 Remote Sensing Retrieval [J]. Urban Surveying, 2024 (4): 69-73
10. [10] Xue Yunfei Research on Predicting Automotive Carbon Dioxide Emissions Based on Machine Learning [J]. Automation and Information Engineering, 2023, 44 (1): 22-26+45
11. [11] Ji Danjun Time Series Analysis and Prediction of Urban Air Quality Index: A Case Study of Yangzhou City [J]. Journal of Wuxi Vocational and Technical College of Commerce, 2024, 24 (1): 59-65
12. [12] Strict Qi, Zhao Wanying, Yu Zhenwei, etc Prediction of respiratory rate in dairy cows using a random forest model based on hyperparameter optimization algorithm [J]. Journal of Agricultural Engineering, 2024, 40 (11): 195-203
13. [13] Zhu Hua, Qiao Yongjin, Dong Guogang Distributed Data Outlier Detection Algorithm Based on CART Decision Tree [J]. Modern Electronic Technology, 2024, 47 (16): 157-162

Funding

Liaoning Province Science and Technology Plan Joint Fund Project (2023-MSLH-256)

About the authors

Zhu Xiaofeng (2000-), male, Han ethnicity, from Anyang, Henan Province, is currently pursuing a master's degree in environmental protection new technologies and key equipment; Zhang Qian (1989-), female, Han ethnicity, from Zhoukou, Henan Province, lecturer, doctoral student, research direction: environmental quality monitoring and analysis; Zhu Xiaoping (1986-), male, Han ethnicity, from Chongqing, associate professor, doctoral student, research direction: environmental quality monitoring and analysis.