判别分析案例(鸢尾花)

特征值函数特征值方差的%累积%正则相关性130.419a99.099.0.9842.293a1.0100.0.476a.分析中使用了前2个典型判别式函数。Wilks的Lambda函数检验Wilks的Lambda卡方dfSig.1到2.025538.9508.0002.77437.3513.000标准化的典型判别式函数系数函数12花萼长-.346.039花萼宽-.525.742花瓣长.846-.386花瓣宽.613.555花瓣宽花瓣长花萼宽花萼长zzzzD0.6130.8460.525-346.01花瓣宽花瓣长花萼宽花萼长zzzzD0.5550.386-0.7420.0392结构矩阵函数12花瓣长.726*.165花萼宽-.121.879*花瓣宽.651.718*花萼长.221.340*判别变量和标准化典型判别式函数之间的汇聚组间相关性按函数内相关性的绝对大小排序的变量。*.每个变量和任意判别式函数间最大的绝对相关性典型判别式函数系数函数12花萼长-.063.007花萼宽-.155.218花瓣长.196-.089花瓣宽.299.271(常量)-2.526-6.987非标准化系数花瓣宽花瓣长花萼宽花萼长zzzzD0.2990.1960.155-063.0526.21花瓣宽花瓣长花萼宽花萼长zzzzD0.2710.089-0.2180.007978.62区域图典则判别函数2-16.0-12.0-8.0-4.0.04.08.012.016.0+---------+---------+---------+---------+---------+---------+---------+---------+16.0+13+I13II13II123II123II1223I12.0++++1223++++I1223II1223II1223II1223II1223I8.0++++12+23++++I1223II1223II1223II1223II1223I4.0++++12+23++++I1223II1223II1223II1223II1223*I.0+++*+12+23+++I12*23II1223II1223II1223II1223I-4.0++++12++23+++I1223II1223II1223II1223II1223I-8.0++++12++23+++I1223II1223II1223II1223II1223I-12.0+++12++23++I1223II1223II1223II1223II1223I-16.0+1223++---------+---------+---------+---------+---------+---------+---------+---------+-16.0-12.0-8.0-4.0.04.08.012.016.0典则判别函数1区域图中使用的符号符号组标签--------------------11刚毛鸢尾花22变色鸢尾花33佛吉尼亚鸢尾花*表示一个组质心分类结果a分类预测组成员合计刚毛鸢尾花变色鸢尾花佛吉尼亚鸢尾花初始计数刚毛鸢尾花500050变色鸢尾花048250佛吉尼亚鸢尾花014950%刚毛鸢尾花100.0.0.0100.0变色鸢尾花.096.04.0100.0佛吉尼亚鸢尾花.02.098.0100.0a.已对初始分组案例中的98.0%个进行了正确分类。