介绍了分位数回归的原理和方法,相比于传统的沉降预测方法,分位数回归以其高精度、计算效率高、高效及稳健的特点,能更好地应用于沉降监测数据的预测,并结合地面沉降实例进行分析和预报,将分位数回归模型预测结果与多项式模型及指数平滑法预测结果进行比较,结果显示分位数回归法预测精度更高。 The principle and method of quantile regression are introduced. Compared with the traditional settlement prediction method, quantile regression can be better applied to the prediction of settlement monitoring data with its high precision, high computational efficiency, high efficiency and robustness , And combined with examples of land subsidence for analysis and forecasting. The prediction results of quantile regression model are compared with those of the polynomial model and exponential smoothing method. The results show that the prediction accuracy of quantile regression method is higher.