随机变量对数化的相关分析(续完)

Correlation analysis of stochastic variables logarithmization

In the computation of hydrological and water resource analysis, in order to obtain a higher correlation coefficient, taking two synchronization series of variables, one variable or two variables, at the same time are taken to be logarithmized, of which the method is considered to be classified as a category of "false correlation". For this reason, in this paper, proceeding from the fundamental mathematic principle, some deriation and anglysis has been made for the correlation of variables logarithmization, and it is shown by the results that after logarithmization, the correlation can not be definitely improved It has been proved by some applicable examples in this paper, showing again that after logarithmization the correlation will come down instead, thus, the attention should be paid in order to avoid the

In the computation of hydrological and water resource analysis, in order to obtain a higher correlation coefficient, taking two synchronization series of variables, one variable or two variables, at the same time are taken to be logarithmized, of which the method is considered to be classified as a category of 'false correlation'. For this reason, in this paper, proceeding from the fundamental mathematic principle, some deriation and anglysis has been made for the correlation of variables logarithmization, and it is shown by the results that after logarithmization, the correlation can not be definitely improved It has been proved by some applicable examples in this paper, showing again that after logarithmization the correlation will come down instead, thus, the attention should be paid in order to avoid the misunderstanding