常规线性回归方法用于建立选厂生产经验模型时的不足及其修正方法

THE SHORTAGE OF THENORMAL LINEAR REGRESSION METHOD WHEN USED IN BUILDING THE EXPERIENCE MODEL FOR THE PRODUCTION OF A MINERAL PROCESSING PLANT AND ITS SOLUTION

由于受现有测量技术的限制，选厂生产经验模型中有影响十分显著的因素无法给予考虑，线性回归的剩余方差较大．常规线性回归方法，由于其优化目标存在缺陷，不适于剩余方差较大的场合，因而需要给予修正．

Because of the constraint of the current technology in measuring,some of the variables of great significance cannot be considered in the experience model for the production of a mineral processing plant, and the residual variance of its linear regression is considerably large. The normal linear regression method is not suitable for the case of large residual variance by reason of the shortage in its optimization objective. Therefore, the method needs to be modified.