数据挖掘实验报告.docx

数据挖掘实验报告.docx

《数据挖掘实验报告.docx》由会员分享，可在线阅读，更多相关《数据挖掘实验报告.docx（14页珍藏版）》请在冰豆网上搜索。

数据挖掘实验报告

《数据挖掘》实验报告

实验序号：

　　　　　　　　　　实验项目名称：

C4.5算法

学　　号

姓　　名

专业、班

12数学金融

实验地点

实验楼5-510

指导教师

潘巍巍

实验时间

2014.12.24

一、实验目的及要求

1：

选择一个数据挖掘标准数据集，采用C4.5算法进行分类，给出分类精度，画出用C4.5算法诱导的树并写出生成的规则集合。

2：

在数据挖掘标准数据集上，实验对比剪枝及未剪枝的树的分类性能。

3：

总结C4.5算法的优缺点

二、实验设备（环境）及要求

电脑WEKA3.6.1

三、实验内容及步骤

（3）数据分类（c4.5算法实现）

1.导入数据

（2）选择C4.5分类器进行分类

结果为

其中分类精度为50%

生成的决策树为

分类规则：

J48prunedtree

------------------

outlook=sunny

|humidity=high:

no（3.0）

|humidity=normal:

yes（2.0）

outlook=overcast:

yes（4.0）

outlook=rainy

|windy=TRUE:

no（2.0）

|windy=FALSE:

yes（3.0）

剪枝后结果为

分类精度变为57.1%性能变好

（1）C4.5算法优缺点

优点：

分类精度高，生成的分类规则比较简单，易于理解。

缺点：

需要多次扫描数据集，比较低效

五、分析及讨论

六、教师评语

签名：

日期：

成绩

《数据挖掘》实验报告

实验序号：

　　　　　　　　　　实验项目名称：

KNN算法

学　　号

姓　　名

专业、班

12数学金融

实验地点

实验楼5-510

指导教师

潘巍巍

实验时间

2014.12.24

一、实验目的及要求

1：

KNN算法的基本思路、步骤。

2：

选择UCI中的5个标准数据集，使用KNN算法在该数据集上计算混淆矩阵。

3：

选择2个数据集，选择不同的k值，k=1,3,5,7,9，对比KNN算法计算结果的差异。

二、实验设备（环境）及要求

电脑WEKA3.6.1

四、实验内容及步骤

1.数据集contact-lenses.arff

Glass.arff

两者的混淆矩阵分别为

（2）两个数据集在K=1,3,5,7,9下结果分别为

Glass：

K=1;

===Summary===

CorrectlyClassifiedInstances15170.5607%

IncorrectlyClassifiedInstances6329.4393%

Kappastatistic0.6005

Meanabsoluteerror0.0897

Rootmeansquarederror0.2852

Relativeabsoluteerror42.3747%

Rootrelativesquarederror87.8627%

TotalNumberofInstances214

===DetailedAccuracyByClass===

TPRateFPRatePrecisionRecallF-MeasureROCAreaClass

0.7860.1670.6960.7860.7380.806buildwindfloat

0.6710.130.7390.6710.7030.765buildwindnon-float

0.2940.0510.3330.2940.3130.59vehicwindfloat

00000?

vehicwindnon-float

0.7690.030.6250.7690.690.895containers

0.7780.0150.70.7780.7370.838tableware

0.7930.0110.920.7930.8520.884headlamps

WeightedAvg.0.7060.1090.7090.7060.7040.792

===ConfusionMatrix===

abcdefg<--classifiedas

55960000|a=buildwindfloat

155140321|b=buildwindnon-float

9350000|c=vehicwindfloat

0000000|d=vehicwindnon-float

02001001|e=containers

0100170|f=tableware

03002123|g=headlamps

K=3;

===Summary===

CorrectlyClassifiedInstances15471.9626%

IncorrectlyClassifiedInstances6028.0374%

Kappastatistic0.6097

Meanabsoluteerror0.0983

Rootmeansquarederror0.2524

Relativeabsoluteerror46.4438%

Rootrelativesquarederror77.7792%

TotalNumberofInstances214

===DetailedAccuracyByClass===

TPRateFPRatePrecisionRecallF-MeasureROCAreaClass

0.8430.2150.6560.8430.7380.865buildwindfloat

0.7110.1380.740.7110.7250.835buildwindnon-float

0.1760.0150.50.1760.2610.672vehicwindfloat

00000?

vehicwindnon-float

0.6150.0150.7270.6150.6670.913containers

0.7780.010.7780.7780.7780.914tableware

0.7930.0110.920.7930.8520.885headlamps

WeightedAvg.0.720.1230.7180.720.7080.847

===ConfusionMatrix===

abcdefg<--classifiedas

591010000|a=buildwindfloat

195420100|b=buildwindnon-float

10430000|c=vehicwindfloat

0000000|d=vehicwindnon-float

0300802|e=containers

0100170|f=tableware

21001223|g=headlamps

K=5;

===Summary===

CorrectlyClassifiedInstances14567.757%

IncorrectlyClassifiedInstances6932.243%

Kappastatistic0.5469

Meanabsoluteerror0.1085

Rootmeansquarederror0.2563

Relativeabsoluteerror51.243%

Rootrelativesquarederror78.9576%

TotalNumberofInstances214

===DetailedAccuracyByClass===

TPRateFPRatePrecisionRecallF-MeasureROCAreaClass

0.8430.2290.6410.8430.7280.867buildwindfloat

0.6840.1740.6840.6840.6840.848buildwindnon-float

00.010000.642vehicwindfloat

00000?

vehicwindnon-float

0.3850.0250.50.3850.4350.952containers

0.6670.010.750.6670.7060.909tableware

0.7930.0160.8850.7930.8360.89headlamps

WeightedAvg.0.6780.1420.6350.6780.6510.853

===ConfusionMatrix===

abcdefg<--classifiedas

591010000|a=buildwindfloat

205210300|b=buildwindnon-float

12500000|c=vehicwindfloat

0000000|d=vehicwindnon-float

0500503|e=containers

0200160|f=tableware

12001223|g=headlamps

K=7;===Summary===

CorrectlyClassifiedInstances13764.0187%

IncorrectlyClassifiedInstances7735.9813%

Kappastatistic0.4948

Meanabsoluteerror0.1147

Rootmeansquarederror0.2557

Relativeabsoluteerror54.1689%

Rootrelativesquarederror78.7876%

TotalNumberofInstances214

===DetailedAccuracyByClass===

TPRateFPRatePrecisionRecallF-MeasureROCAreaClass

0.8290.2710.5980.8290.6950.876buildwindfloat

0.6050.1810.6480.6050.6260.852buildwindnon-float

0.0590.0050.50.0590.1050.71vehicwindfloat

00000?

vehicwindnon-float

0.3080.030.40.3080.3480.939containers

0.5560.0150.6250.5560.5880.976tableware

0.7930.0160.8850.7930.8360.89headlamps

WeightedAvg.0.640.1580.6360.640.6170.864

===ConfusionMatrix===

abcdefg<--classifiedas

581110000|a=buildwindfloat

264600400|b=buildwindnon-float

11510000|c=vehicwindfloat

0000000|d=vehicwindnon-float

0500413|e=containers

1200150|f=tableware

12001223|g=headlamps

K=9;

===Summary===

CorrectlyClassifiedInstances13563.0841%

IncorrectlyClassifiedInstances7936.9159%

Kappastatistic0.4782

Meanabsoluteerror0.1196

Rootmeansquarederror0.2581

Relativeabsoluteerror56.4924%

Rootrelativesquarederror79.5178%

TotalNumberofInstances214

===DetailedAccuracyByClass===

TPRateFPRatePrecisionRecallF-MeasureROCAreaClass

0.8290.2780.5920.8290.690.881buildwindfloat

0.6450.1740.6710.6450.6580.853buildwindnon-float

00.0050000.694vehicwindfloat

00000?

vehicwindnon-float

0.2310.030.3330.2310.2730.933containers

0.2220.0150.40.2220.2860.964tableware

0.7930.0270.8210.7930.8070.888headlamps

WeightedAvg.0.6310.1590.580.6310.5970.864

===ConfusionMatrix===

abcdefg<--classifiedas

581110000|a=buildwindfloat

234900310|b=buildwindnon-float

13400000|c=vehicwindfloat

0000000|d=vehicwindnon-float

0600304|e=containers

3100221|f=tableware

12001223|g=headlamps

contact-lenses:

K=1;

===Summary===

CorrectlyClassifiedInstances1979.1667%

IncorrectlyClassifiedInstances520.8333%

Kappastatistic0.6262

Meanabsoluteerror0.2262

Rootmeansquarederror0.3165

Relativeabsoluteerror59.8856%

Rootrelativesquarederror72.4707%

TotalNumberofInstances24

===DetailedAccuracyByClass===

TPRateFPRatePrecisionRecallF-MeasureROCAreaClass

0.80.0530.80.80.80.958soft

0.750.10.60.750.6670.925hard

0.80.2220.8570.80.8280.896none

WeightedAvg.0.7920.1670.8020.7920.7950.914

===ConfusionMatrix===

abc<--classifiedas

401|a=soft

031|b=hard

1212|c=none

K=3;

===Summary===

CorrectlyClassifiedInstances1979.1667%

IncorrectlyClassifiedInstances520.8333%

Kappastatistic0.6262

Meanabsoluteerror0.2262

Rootmeansquarederror0.3165

Relativeabsoluteerror59.8856%

Rootrelativesquarederror72.4707%

TotalNumberofInstances24

===DetailedAccuracyByClass===

TPRateFPRatePrecisionRecallF-MeasureROCAreaClass

0.80.0530.80.80.80.958soft

0.750.10.60.750.6670.925hard

0.80.2220.8570.80.8280.896none

WeightedAvg.0.7920.1670.8020.7920.7950.914

===ConfusionMatrix===

abc<--classifiedas

401|a=soft

031|b=hard

1212|c=none

K=5;

===Summary===

CorrectlyClassifiedInstances1666.6667%

IncorrectlyClassifiedInstances833.3333%

Kappastatistic0.3356

Meanabsoluteerror0.2793

Rootmeansquarederror0.3624

Relativeabsoluteerror73.9227%

Rootrelativesquarederror82.9705%

TotalNumberofInstances24

===DetailedAccuracyByClass===

TPRateFPRatePrecisionRecallF-MeasureROCAreaClass

0.60.0530.750.60.6670.947soft

0.250.10.3330.250.2860.856hard

0.80.5560.7060.80.750.859none

WeightedAvg.0.6670.3750.6530.6670.6550.877

===ConfusionMatrix===

abc<--classifiedas

302|a=soft

013|b=hard

K=7;===Summary===

CorrectlyClassifiedInstances1458.3333%

IncorrectlyClassifiedInstances1041.6667%

Kappastatistic-0.0619

Meanabsoluteerror0.3188

Rootmeansquarederror0.387

Relativeabsoluteerror84.3959%

Rootrelativesquarederror88.61%

TotalNumberofInstances24

===DetailedAccuracyByClass===

TPRateFPRatePrecisionRecallF-MeasureROCAreaClass

00.0530000.947soft

000000.831hard

0.93310.6090.9330.7370.807none

WeightedAvg.0.5830.6360.380.5830.4610.841

===ConfusionMatrix===

abc<--classifiedas

005|a=soft

004|b=hard

1014|c=none

K=9;

===Summary===

CorrectlyClassifiedInstances1458.3333%

IncorrectlyClassifiedInstances1041.6667%

Kappastatistic-0.0619

Meanabsoluteerror0.3188

Rootmeansquarederror0.387

Relativeabsoluteerror84.3959%

Rootrelativesquarederror88.61%

TotalNumberofInstances24

===DetailedAccuracyByClass===

TPRateFPRatePrecisionRecallF-MeasureROCAreaClass

00.0530000.947soft

000000.831hard

0.93310.6090.9330.7370.807none

WeightedAvg.0.5830.6360.380.5830.4610.841

===ConfusionMatrix===

abc<--classifiedas

005|a=soft

004|b=hard

1014|c=none

可以看出第一个数据集在K=3时分类精度最高