还剩5页未读,继续阅读
文本内容:
验报告实验名称SPSS的聚类分析学号专业班级课程名称统计分析SPSS软件实验室
一、实验目的:指导教师掌握层次聚类分析和K-Means聚类分析的基本思想和具体,并能够对分析结果进行解释
二、实验题目:、现要对一个班同学的语文水平进行聚类,拟聚为三类,聚类依据是两次语文考试的成绩数据如下表所示试
1.用系统聚类法和.均值法进行聚类分析K人名第一次语文成绩第二次语文成绩张三9998王五8889赵四7980小杨8978蓝天7578小白6065李之7987马武7576郭炎6056刘小100100
三、实验步骤(最好有截图):L先打开常用软件里的SPSS
11.5for Windows,exe,在Variable View中根据题目输入相关数据,如下图所示BUntitled-SPSS DataEditorI%©■ip昌同卜出国耻闻唯阖勖雕JU」File EditView Data Transform Analyze Graphs UtilitiesWindow Help—■■—■■r■■■.Name TypeWidth DecimalsLabel ValuesMissing ColumnsAlign Measure人名_______1String J80None None8Left Nominal第一次_2Numeric82None None8Right Scale语第二孀_3Numeric82None None8Right Scale_
42.在Data View中先输入数据,结果如下图所示国Untitled-SPSS DataEditorFile Ed:L tVi ewDataTransformAnalyzeGraphsUtilities WindovY Help以旧|昌|曳口|阖灯广|□〔工濠|寓I L|j]IC10人名|第一次语|第二枚语var varvar张三199,
0098.002王五88,
0089.003赵四79,
0080.004小杨89,
0078.005蓝天75,
0078.006小白60,
0065.007李之79,
0087.008马武75,
0076.009郭炎60,
0056.00刘小口口.口口
101100.00_J
3.首先试用系统聚类法对相关数据进行聚类
4.选择菜单【Analyze】f[Classify]f[Hierarchical Cluster],然后选择参与层次聚类分析的变量两次语文考试的成绩到【Variable s】框中,再选择一个字符型变量“人名”作为标记变量至U【LabelCases by]框中
5.按“Plots”后进行选择Hierarchical ClusterAnalysis:Plots
7.按“Method”后进行选择▽[DendrogramContinue「Icicle0All clustersSpecified rangeofCancelclustersHelpStart[1Stop IBy|1NoneOrientation0Vertical Horizontal
6.按“Statistics”后进行选择Hierarchical ClusterAnalysis:Statistics*Agglomeration scheduleContinue一Proximity matrixCancelClusterMembershipHelp0None Singlesolution clusters「Range ofsolutionsthrough IFromclusters ClusterMethod Between-groups linkageContinueMeasureCancel*Interval SquaredEuclidean distanceHelpM qRoot CountsChi-square measure二binary SquaredEuclidean distancePresent1Absent0Transform ValuesTransform MeasuresAbsolute valuesStandardizeNoneF Changesign二]_Rescale to0-1rangeG ByvariableC Bycase
8.对第一个表格进行保存,并且命名为“语文水平.sav”,同时保存输出结果
四、实验结果及分析(最好有截图)第一题
1.首先试用系统聚类法对相关数据进行聚类4ClusterCase ProcessingSummary3也CasesValid MissingTotalN PercentN PercentN Percent
10100.00,
010100.0,a SquaredEuclidean Distanceusedb.Average Linkage Between GroupsAverage LinkageBetweenGroupsAgglomeration ScheduleStageCluster FirstClusterCombined AppearsStageCluster1Cluster2Coefficients Cluster1Cluster2Next Stage
1584.
00000321105.00000833526,00007146981,00000952785,000006624151,500507723174,778638812717,
8332799161474.250840Vertical IcicleCase96853472郭小q就起小至王张炎自武天四杨之五—Number ofclusters1X X X X X X X X X X X X X X X X X X X2X X X XXX X X X X X X X X X X X X3X X X XXX X X X X X X X X X X X4X XX XXX XX XXXXXXXX5XXX XXXXXXXXXXXX6XXXXXXXXXXXXXX7XXXXXXXXXXXXX8XXXXX XXXXXXX9XXXXX XXXXXXDendrogramCLUSTER ANALYSIS******♦♦♦♦♦♦HIERARCHICALDendrogram usingAverageLinkageBetween GroupsRescaledDistance ClusterCombineC A10152025Label Num-+赵天5王武8李四小五3张之2刘杨1小三郭小4白1炎10蓝6马
92.K-均值法进行聚类分析后的输出结果4Quick ClusterInitialCluster CentersCluster123第T欠语
100.
0060.
0079.00第二次语
100.
0056.
0087.00Iteration History3Change inCluster CentersIteration
12311.
1184.
5005.9562,000,000,
0003.Convergence achieveddue tono orsmall changein clustercenters.The maximumabsolutecoordinate changefor anycenter is,
000.The currentiteration is2,The minimumdistance betweeninitialcenters is
24.
698.Final ClusterCentersCluster123第T欠港
99.
5060.
0080.83第二次语
99.
0060.
5081.33ANOVACluster ErrorMean Square dfMeanSquaredf FSig.第T欠港781,533227,619728,297,000第二次语744,133226,548728,030,000The Ftests shouldbe usedonly fordescriptive purposesbecause theclusters havebeen chosentomaximize thedifferences amongcases indifferent clusters.The observedsignificance levelsare notcorrectedfor thisand thuscannot beinterpreted astests ofthe hypothesisthat thecluster meansareequal.Number ofCases ineach ClusterCluster
12.
00022.
00036.000Valid Missing10,000,000。