我国财政支出的聚类分析1

《我国财政支出的聚类分析1》由会员上传分享，免费在线阅读，更多相关内容在工程资料-天天文库。

1、徐金华20091910228我国地区产业的就职人员结构聚类分析金融与统计学院一、数据:（万人）第一产第二产第三产业业业65.7263.8925.677.7209.4220.21483.61214.71201.4635.8419.4544.5558193.3391.2694.4559.6936516.6239.7428.4781343.9562.647.6347.5534.2896.92030.91608.4659.31795.41370.41579.61041.7106&4638.6775.7754.6882.3635.5726.41994.41741.2

2、1714.12764.91674.71509.2990.1814.71219.71876.4814.91216.31536.71922.72183.91561.2582.1719.3226.149.3156.1655.7519.5703.22158.31110.11676.91210.5268.3862.31672.7353.4704.192.21&25&7877.6432.5609.3739.3204.8462.5122.662.8100.1130.884.7113425.8116.3287.2就业人数京津北西古宁林江海苏江徽建西东南北南东西南庆川州南藏

3、西肃海夏疆区蒙龙地北天河山内辽吉黑上江浙安福江山河湖湖广广海重四贵云西陕廿青宁新二、实验步骤：1将数据导入spss建立项Fl文件，点击analyze-^classify-►hierarchicalcluster,将变量选入variable栏中，地区选入labelcaseby栏中；2、点击statistic按钮，选择输出agglomerationschedule、proximitymatrix两项，聚类范围为3至06,单击continue返回对话框；3、点击plots按钮，选择dendrogram树状图，单击continue返回；4、点击method,数值标

4、准化为Z-scores,聚类方法为between-groupslinkage,选择距离测度为squarededuclideandistance,返回主对话框;5、点击save,rangeofsolution范围为3到6,返冋主对话框，点击ok提交系统执行。在输出窗口屮显示结果清单。三、实验结果：表1ClusterMembershipCase6Clusters5Clusters4Clusters3Clusters北京11112:天津11113:河北22214:山西11115:内蒙古11116:辽宁11117:吉林11118:黑龙江11119上海111110:

5、江苏3332X:浙江333212:安徽222113:福建111114:江西111115：

6、1

7、东444316:河南444317:湖北222148:湖南222119:广东553220:广西622121:海南111122:ffi庆111123:四川444324:贵州622125:云南622126:西藏111127:陕西11112&甘肃111129:青海111130:宁夏111134:新疆1111CaseProcessingSummary3CasesValidMissingTotalNPercentNPercentNPercent31100.0%0.0%3110

8、0.0%a.SquaredEuclideanDistanceused表3AgglomerationScheduleStageClusterCombinedCoefficientsStageClusterFirstAppearsNextStageCluster1Cluster2Cluster1Cluster212930.002003257.01400632629.01901442126.0440385827.0480076528.0962012748.09605158221.129042591422.143001310312.1610019112025.16

9、9001812531.1806015131314.183091414613.225013201545.38971220161011.45900271719.5420023182024.5631102619318.572100222046.8511514232115231.1460024223171.1511902623141.2611720252415161.5372102925121.80623828263201.9142218282710192.9261602928134.7152526302910154.9502724303011012.898282

10、90四、实验结果分析：（1）表1现实的是将样本分为3类、4类、5类