论文部分内容阅读
三、逐步多类判别分析变量增多,判别能力变好;但当变量增至一定数量时,若再增多,判别能力常会降低。如由于自变量间的不独立性,可能导致计算S~(-1)时精度下降或出现困难(如病态、甚至退化),使最后建立的判别函数不稳定,错判样品个数增加,判别能力差。当变量增多,建立判别函数也需更多的计算工作量,在应用时也不方便。因此在可供判别分类的自变量中选出重要变量非常
Third, gradually increasing the number of discriminant analysis of variables, discriminatory ability better; but when the variable increases to a certain number, if it increases, discriminatory ability often decreases. For example, the indifference between independent variables may lead to the decrease of accuracy or difficulty (such as pathological or even degradation) in the calculation of S ~ (-1), the instability of the final established discriminant function and the increase of the number of wrongly-judged samples. poor ability. When the variable increases, the establishment of discriminant function also needs more computational workload, which is inconvenient in application. Therefore, it is very important to choose the important variables among the variables that can be classified