判别分析三种方法

《判别分析三种方法》由会员上传分享，免费在线阅读，更多相关内容在行业资料-天天文库。

1、作业一：为研究1991年中国城镇居民月平均收入状况，按标准化欧氏平方距离、离差平方和聚类方法将30个省、市、自治区．分为两种类型。试建立判别函数，判定广东、西藏分别属于哪个收入类型。判别指标及原始数据见表9-4。1991年30个省、市、自治区城镇居民月平均收人数据表单位：元／人x1：人均生活费收入x6：人均各种奖金、超额工资(国有+集体)x2：人均国有经济单位职工工资x7：人均各种津贴(国有+集体)x3：人均来源于国有经济单位标准工资x8：人均从工作单位得到的其他收入x4：人均集体所有制工资收入x9：个体劳动者收入x5：人均集体所有制职工标准工资样

2、品序号地区x1x2x3x4x5x6x7x8x9类序号G11北京170.03110.259.768.384.4926.8016.4411.90.412天津141.5582.5850.9813.49.3321.3012.369.211.053河北119.4083.3353.3911.07.5217.3011.7912.00.704上海194.53107.860.2415.68.8831.0021.0111.80.165山东130.4686.2152.3015.910.520.6l12.149.610.476湖北119.2985.4153.0213.18

3、.4413.8716.478.380.517广西134.4698.6148.188.904.3421.4926.1213.64.568海南143.7999.9745.606.301.5618.6729.4911.83.829四川128.0574.9650.1313.99.6216.1410.1814.51.2110云南127.4193.5450.5710.55.8719.4121.2012.60.9011新疆122.96101.469.706.303.8611.3018.965.624.62G21山西102.4971.7247.729.426.96

4、13.127.96.660.612内蒙古106.1476.2746.199.656.279.65520.1O6.970.963吉林104.9372.9944.6013.79.019.43520.616.651.684黑龙江103.3462.9942.9511.17.4l8.34210.196.452.685江西98.08969.4543.0411.47.9510.5916.507.691.086河南104.1272.2347.319.486.4313.1410.438.301.117贵州108.4980.7947.526.063.4213.6916

5、.538.372.858陕西113.9975.650.885.213.8612.949.4926.771.279甘肃114.0684.3152.787.815.4410.8216.433.791.1910青海108.8080.4150.457.274.078.37118.985.950.8311宁夏115.9688.2l51.858.815.6313.9522.654.750.9711待判1广东211.30114.041.4433.211.248.7230.7714.911.12西藏175.93163.857.894.223.3717.8182.3

6、215.70.00一、距离判别法解：变量个数p＝9，两类总体各有11个样品，即n1＝n2＝11，有2个待判样品，假定两总体协差阵相等。由spss可计算出：协方差和平均值汇聚的组内矩阵x1x2x3x4x5x6x7x8x9相关性x11.000.217.299.045-.054.688.212.121-.245x2.2171.000.102-.234-.211.136-.052.116.154x3.299.1021.000-.296-.062.091-.017-.607-.034x4.045-.234-.2961.000.762-.172-.297.10

7、3-.554x5-.054-.211-.062.7621.000-.156-.342.022-.654x6.688.136.091-.172-.1561.000.235.384-.098x7.212-.052-.017-.297-.342.2351.000-.040.424x8.121.116-.607.103.022.384-.0401.000-.071x9-.245.154-.034-.554-.654-.098.424-.0711.000组统计量VAR00010均值标准差有效的N（列表状态）未加权的已加权的1.00x1139.266423.3

8、51251111.000x293.091811.388291111.000x353.98826.805301111.000