论文部分内容阅读
在城市水资源承载能力研究中,偏最小二乘回归方法能有效地处理自变量间多重线性相关性问题,但不能较好地处理因变量与自变量间复杂的非线性问题.投影寻踪神经网络耦合模型是处理非线性问题的有力工具,而且神经网络投影寻踪耦合模型稳健性高,但不能较好地处理自变量间多重线性相关性问题.本文把这两种方法结合在一起,建立了基于偏最小二乘回归的神经网络投影寻踪耦合模型,对城市水资源承载能力进行了预测,并取得了满意效果.
In the study of urban water resources carrying capacity, partial least-squares regression method can effectively deal with the problem of multiple linear correlation between independent variables, but can not deal with the complex nonlinear problem between dependent variable and independent variable. The network coupling model is a powerful tool to deal with the nonlinear problems, and neural network projection pursuit coupled model is robust, but it can not deal with the multiple linear correlation problems well.We combine these two methods to establish Based on partial least squares regression neural network projection pursuit coupled model, the urban water resources carrying capacity is predicted, and the satisfactory results are obtained.