高斯过程回归模型在河流水温模拟中的应用

Application of Gaussian process regression models in river water temperature modelling

通过收集密苏里河三个站点的多年实测水温、气温数据,建立了水温-气温的高斯过程回归模型.和传统的水温回归模型（线性、非线性及随机回归模型）相比,高斯过程回归模型的精度较高,各研究站点的相关系数值均较高（0.966 4-0.989 7）,均方根误差值相对较小（1.978 4-1.495 0）.高斯过程回归作为一种较为先进的机器学习方法,和其他机器学习方法相比,其突出的优点在于结合了许多机器学习任务,包括模型训练、不确定性分析及超参数估计等,该方法可以有效地应用于天然河流站点水温的预测.

By collecting long-term observed water and air temperature data from three stations in the Missouri River,water-air temperature relationships were developed with Gaussian process regression models.Gaussian process regression models outperform the traditional regression methods（linear,non-linear and stochastic regression models） with larger coefficient of correlations（0.966 4-0.989 7） and smaller root mean square errors（1.978 4-1.495 0） for the three studied stations.Gaussian process regression is an advanced machine learning method.Compared to other machine learning methods,the advantage of Gaussian process regression is that it combines several machine learning tasks,including model training,uncertainty estimation and hyperparameter estimation,and it can be used to effectively predict water temperature in the natural river sections.