Prediction of air quality based on Gradient Boosting Machine Method
Yuelai Su
Abstract
With the rapid development of China's economy, the degree of industrialization is gradually deepened, therefore leading to environmental pollution problems. Air is the material basis on which human beings live. Beijing is the capital of China and the national economic, political and cultural center, Beijing's air quality is an important indicator to measure whether the city is livable or not, while PM2.5 has also become an important standard to measure and monitor the air quality of Beijing. In today's era of big data, the use of efficient computing software to conduct data analysis and prediction has become a trend of future environmental detection and data analysis, which can effectively monitor the urban air environment. In this paper, two methods, Light Gradient Boosting Machine (Light GBM) and eXtreme Gradient Boosting (XGB) were used to extract and predict the characteristics of the air monitoring data in Beijing, and the prediction accuracy and operation time of the two methods were evaluated. Finally, conclusion was drawn that the accuracy and operation efficiency of Light GBM was much higher than that of XGB was reached.