Car Evaluation Analysis
汽车数据R语言机器学习分析
- title: “Car Evaluation Analysis”
- author: “Suraj Vidyadaran”
- date: “Sunday, February 21, 2016”
- output: md_document
对汽车数据使用17种分类算法进行数据分析,对代码进行实践应用和内容翻译
- Load the data 读取数据
- Exploratory Data Analysis 探索数据
- Classification Analysis 分类
Load the data 读取数据
buying 价格 maint 大小 doors 门 persons 载人 lug_boot 轮子 safety 安全 class 等级
1 | car_eval <- read.csv("car.data.txt", header=FALSE) |
## buying maint doors persons lug_boot safety class
## 1 vhigh vhigh 2 2 small low unacc
## 2 vhigh vhigh 2 2 small med unacc
## 3 vhigh vhigh 2 2 small high unacc
## 4 vhigh vhigh 2 2 med low unacc
## 5 vhigh vhigh 2 2 med med unacc
## 6 vhigh vhigh 2 2 med high unaccExploratory Data Analysis 探索数据
# summary(car_eval)
str(car_eval)
'data.frame': 1728 obs. of 7 variables:
$ buying : Factor w/ 4 levels "high","low","med",…: 4 4 4 4 4 4 4 4 4 4 …
$ maint : Factor w/ 4 levels "high","low","med",…: 4 4 4 4 4 4 4 4 4 4 …
$ doors : Factor w/ 4 levels "2","3","4","5more": 1 1 1 1 1 1 1 1 1 1 …
$ persons : Factor w/ 3 levels "2","4","more": 1 1 1 1 1 1 1 1 1 2 …
$ lug_boot: Factor w/ 3 levels "big","med","small": 3 3 3 2 2 2 1 1 1 3 …
$ safety : Factor w/ 3 levels "high","low","med": 2 3 1 2 3 1 2 3 1 2 …
$ class : Factor w/ 4 levels "acc","good","unacc",…: 3 3 3 3 3 3 3 3 3 3 …
Classification Analysis 分类
Linear Classification 线性分类
1 Logistic Regression 逻辑回归
library(VGAM) # 广义线性模型
#Build the model
model1<-vglm(class~buying+maint+doors+persons+lug_boot+safety,family = "multinomial",data=car_eval)
#Summarize the model
summary(model1)
#Predict using the model
x<-car_eval[,1:6]
y<-car_eval[,7]
probability<-predict(model1,x,type="response") # 相应就是二元
car_evalpred_log_reg<-apply(probability,1,which.max)
car_evalpred_log_reg[which(car_evalpred_log_reg=="1")]<-levels(car_evalclass)[1]
car_evalpred_log_reg=="2")]<-levels(car_evalpred_log_reg[which(car_evalpred_log_reg=="3")]<-levels(car_evalclass)[3]
car_evalpred_log_reg=="4")]<-levels(car_eval$class)[4]
#Accuracy of the model
mtab<-table(car_evalclass)
library(caret)
confusionMatrix(mtab)
Confusion Matrix and Statistics
acc good unacc vgood
acc 340 5 47 2
good 8 59 1 0
unacc 31 0 1162 0
vgood 5 5 0 63
Overall Statistics
Accuracy : 0.9398
95% CI : (0.9275, 0.9506)
No Information Rate : 0.7002
P-Value [Acc > NIR] : < 2.2e-16
Kappa : 0.8699
Mcnemar's Test P-Value : NA
Statistics by Class:
Class: acc Class: good Class: unacc Class: vgood
Sensitivity 0.8854 0.85507 0.9603 0.96923
Specificity 0.9598 0.99458 0.9402 0.99399
Pos Pred Value 0.8629 0.86765 0.9740 0.86301
Neg Pred Value 0.9670 0.99398 0.9103 0.99879
Prevalence 0.2222 0.03993 0.7002 0.03762
Detection Rate 0.1968 0.03414 0.6725 0.03646
Detection Prevalence 0.2280 0.03935 0.6904 0.04225
Balanced Accuracy 0.9226 0.92482 0.9502 0.98161
2 Linear Discriminant Analysis 线性判别
library(caret)
library(MASS)
#Build the model
model2<-lda(class~buying+maint+doors+persons+lug_boot+safety,data=car_eval)
#Summarize the model
summary(model2)
#Predict using the model
car_evalpred_lda<-predict(model2,x)class
#Accuracy of the model
mtab<-table(car_evalclass)
confusionMatrix(mtab)
Confusion Matrix and Statistics
acc good unacc vgood
acc 361 44 70 27
good 10 23 2 3
unacc 13 0 1138 0
vgood 0 2 0 35
Overall Statistics
Accuracy : 0.901
95% CI : (0.886, 0.9147)
No Information Rate : 0.7002
P-Value [Acc > NIR] : < 2.2e-16
Kappa : 0.7883
Mcnemar's Test P-Value : NA
Statistics by Class:
Class: acc Class: good Class: unacc Class: vgood
Sensitivity 0.9401 0.33333 0.9405 0.53846
Specificity 0.8951 0.99096 0.9749 0.99880
Pos Pred Value 0.7191 0.60526 0.9887 0.94595
Neg Pred Value 0.9812 0.97278 0.8752 0.98226
Prevalence 0.2222 0.03993 0.7002 0.03762
Detection Rate 0.2089 0.01331 0.6586 0.02025
Detection Prevalence 0.2905 0.02199 0.6661 0.02141
Balanced Accuracy 0.9176 0.66215 0.9577 0.76863
Non-Linear Classification 非线性分类
3 Mixture Discriminant Analysis 混合判别
library(mda)
#Build the model
model3<-mda(class~buying+maint+doors+persons+lug_boot+safety,data=car_eval)
#Summarize the model
summary(model3)
#Predict using the model
car_eval$pred_mda<-predict(model3,x)
#Accuracy of the model
mtab<-table(car_evalclass)
confusionMatrix(mtab)
Confusion Matrix and Statistics
acc good unacc vgood
acc 353 13 92 16
good 20 51 3 0
unacc 5 0 1115 0
vgood 6 5 0 49
Overall Statistics
Accuracy : 0.9074
95% CI : (0.8928, 0.9207)
No Information Rate : 0.7002
P-Value [Acc > NIR] : < 2.2e-16
Kappa : 0.808
Mcnemar's Test P-Value : NA
Statistics by Class:
Class: acc Class: good Class: unacc Class: vgood
Sensitivity 0.9193 0.73913 0.9215 0.75385
Specificity 0.9100 0.98614 0.9903 0.99339
Pos Pred Value 0.7447 0.68919 0.9955 0.81667
Neg Pred Value 0.9753 0.98912 0.8438 0.99041
Prevalence 0.2222 0.03993 0.7002 0.03762
Detection Rate 0.2043 0.02951 0.6453 0.02836
Detection Prevalence 0.2743 0.04282 0.6481 0.03472
Balanced Accuracy 0.9146 0.86263 0.9559 0.87362
4 Quadratic Discriminant Analysis 二次判别
library(MASS)
#Build the model
model4<-qda(class~buying+maint+doors+persons+lug_boot+safety,data=car_eval)
#Summarize the model
summary(model4)
#Predict using the model
car_eval$pred_qda<-predict(model4,x)
#Accuracy of the model
mtab<-table(car_evalclass)
confusionMatrix(mtab)
5 Regularized Discriminant Analysis 正则约束判别
library(klaR)
#Build the model
model5<-rda(class~buying+maint+doors+persons+lug_boot+safety,data=car_eval,gamma = 0.05,lambda = 0.01)
#Summarize the model
summary(model5)
#Predict using the model
car_evalpred_rda<-predict(model5,x)class
#Accuracy of the model
mtab<-table(car_evalclass)
confusionMatrix(mtab)
Confusion Matrix and Statistics
acc good unacc vgood
acc 317 0 134 0
good 45 63 4 0
unacc 0 0 1072 0
vgood 22 6 0 65
Overall Statistics
Accuracy : 0.8779
95% CI : (0.8615, 0.893)
No Information Rate : 0.7002
P-Value [Acc > NIR] : < 2.2e-16
Kappa : 0.7572
Mcnemar's Test P-Value : NA
Statistics by Class:
Class: acc Class: good Class: unacc Class: vgood
Sensitivity 0.8255 0.91304 0.8860 1.00000
Specificity 0.9003 0.97046 1.0000 0.98316
Pos Pred Value 0.7029 0.56250 1.0000 0.69892
Neg Pred Value 0.9475 0.99629 0.7896 1.00000
Prevalence 0.2222 0.03993 0.7002 0.03762
Detection Rate 0.1834 0.03646 0.6204 0.03762
Detection Prevalence 0.2610 0.06481 0.6204 0.05382
Balanced Accuracy 0.8629 0.94175 0.9430 0.99158
6 Neural Network 神经网络
library(nnet)
#Build the model
model6<-nnet(class~buying+maint+doors+persons+lug_boot+safety,data=car_eval,size = 4,decay = 0.0001,maxit = 500)
# weights: 84
initial value 4315.884207
iter 10 value 494.299222
iter 20 value 271.980488
iter 30 value 201.767107
iter 40 value 170.953147
iter 50 value 110.980954
iter 60 value 56.771482
iter 70 value 41.572251
iter 80 value 34.479180
iter 90 value 30.761660
iter 100 value 29.156225
iter 110 value 28.361403
iter 120 value 27.944054
iter 130 value 27.553794
iter 140 value 26.698026
iter 150 value 25.411406
iter 160 value 24.996274
iter 170 value 24.851429
iter 180 value 24.830112
iter 190 value 24.735811
iter 200 value 24.426379
iter 210 value 23.802677
iter 220 value 23.478896
iter 230 value 23.307991
iter 240 value 23.164112
iter 250 value 23.047281
iter 260 value 22.863117
iter 270 value 22.802107
iter 280 value 22.739661
iter 290 value 22.689016
iter 300 value 22.625042
iter 310 value 22.595026
iter 320 value 22.581459
iter 330 value 22.573767
iter 340 value 22.566224
iter 350 value 22.565535
iter 360 value 22.564483
iter 370 value 22.560267
iter 380 value 22.557433
iter 390 value 22.555073
iter 400 value 22.552698
iter 410 value 22.550126
iter 420 value 22.548059
iter 430 value 22.546974
iter 440 value 22.546031
iter 450 value 22.545549
iter 460 value 22.545069
iter 470 value 22.544765
iter 480 value 22.544515
iter 490 value 22.544258
iter 500 value 22.544140
final value 22.544140
stopped after 500 iterations
#Summarize the model
summary(model6)
#Predict using the model
car_eval$pred_nnet<-predict(model6,x,type="class")
#Accuracy of the model
mtab<-table(car_evalclass)
confusionMatrix(mtab)
Confusion Matrix and Statistics
acc good unacc vgood
acc 384 0 5 0
good 0 69 0 0
unacc 0 0 1205 0
vgood 0 0 0 65
Overall Statistics
Accuracy : 0.9971
95% CI : (0.9933, 0.9991)
No Information Rate : 0.7002
P-Value [Acc > NIR] : < 2.2e-16
Kappa : 0.9937
Mcnemar's Test P-Value : NA
Statistics by Class:
Class: acc Class: good Class: unacc Class: vgood
Sensitivity 1.0000 1.00000 0.9959 1.00000
Specificity 0.9963 1.00000 1.0000 1.00000
Pos Pred Value 0.9871 1.00000 1.0000 1.00000
Neg Pred Value 1.0000 1.00000 0.9904 1.00000
Prevalence 0.2222 0.03993 0.7002 0.03762
Detection Rate 0.2222 0.03993 0.6973 0.03762
Detection Prevalence 0.2251 0.03993 0.6973 0.03762
Balanced Accuracy 0.9981 1.00000 0.9979 1.00000
7 Flexible Discriminant Analysis 弹性判别
library(mda)
#Build the model
model7<-fda(class~buying+maint+doors+persons+lug_boot+safety,data=car_eval)
#Summarize the model
summary(model7)
#Predict using the model
car_eval$pred_fda<-predict(model7,x,type="class")
#Accuracy of the model
mtab<-table(car_evalclass)
confusionMatrix(mtab)
Confusion Matrix and Statistics
acc good unacc vgood
acc 361 44 72 27
good 10 23 2 3
unacc 13 0 1136 0
vgood 0 2 0 35
Overall Statistics
Accuracy : 0.8999
95% CI : (0.8848, 0.9136)
No Information Rate : 0.7002
P-Value [Acc > NIR] : < 2.2e-16
Kappa : 0.786
Mcnemar's Test P-Value : NA
Statistics by Class:
Class: acc Class: good Class: unacc Class: vgood
Sensitivity 0.9401 0.33333 0.9388 0.53846
Specificity 0.8936 0.99096 0.9749 0.99880
Pos Pred Value 0.7163 0.60526 0.9887 0.94595
Neg Pred Value 0.9812 0.97278 0.8722 0.98226
Prevalence 0.2222 0.03993 0.7002 0.03762
Detection Rate 0.2089 0.01331 0.6574 0.02025
Detection Prevalence 0.2917 0.02199 0.6649 0.02141
Balanced Accuracy 0.9169 0.66215 0.9569 0.76863
8 Support Vector Machine 支持向量机
library(kernlab)
#Build the model
model8<-ksvm(class~buying+maint+doors+persons+lug_boot+safety,data=car_eval)
#Summarize the model
summary(model8)
#Predict using the model
car_eval$pred_svm<-predict(model8,x,type="response")
#Accuracy of the model
mtab<-table(car_evalclass)
confusionMatrix(mtab)
Confusion Matrix and Statistics
acc good unacc vgood
acc 375 0 33 0
good 8 60 3 0
unacc 0 0 1174 0
vgood 1 9 0 65
Overall Statistics
Accuracy : 0.9688
95% CI : (0.9594, 0.9764)
No Information Rate : 0.7002
P-Value [Acc > NIR] : < 2.2e-16
Kappa : 0.9333
Mcnemar's Test P-Value : NA
Statistics by Class:
Class: acc Class: good Class: unacc Class: vgood
Sensitivity 0.9766 0.86957 0.9702 1.00000
Specificity 0.9754 0.99337 1.0000 0.99399
Pos Pred Value 0.9191 0.84507 1.0000 0.86667
Neg Pred Value 0.9932 0.99457 0.9350 1.00000
Prevalence 0.2222 0.03993 0.7002 0.03762
Detection Rate 0.2170 0.03472 0.6794 0.03762
Detection Prevalence 0.2361 0.04109 0.6794 0.04340
Balanced Accuracy 0.9760 0.93147 0.9851 0.99699
9 k-Nearest Neighbors k近邻
library(caret)
#Build the model
model9<-knn3(class~buying+maint+doors+persons+lug_boot+safety,data=car_eval,k=5)
#Summarize the model
summary(model9)
#Predict using the model
car_eval$pred_knn<-predict(model9,x,type="class")
#Accuracy of the model
mtab<-table(car_evalclass)
confusionMatrix(mtab)
Confusion Matrix and Statistics
acc good unacc vgood
acc 332 47 11 19
good 6 13 0 4
unacc 44 5 1199 1
vgood 2 4 0 41
Overall Statistics
Accuracy : 0.9172
95% CI : (0.9032, 0.9298)
No Information Rate : 0.7002
P-Value [Acc > NIR] : < 2.2e-16
Kappa : 0.8118
Mcnemar's Test P-Value : 2.234e-13
Statistics by Class:
Class: acc Class: good Class: unacc Class: vgood
Sensitivity 0.8646 0.188406 0.9909 0.63077
Specificity 0.9427 0.993972 0.9035 0.99639
Pos Pred Value 0.8117 0.565217 0.9600 0.87234
Neg Pred Value 0.9606 0.967155 0.9770 0.98572
Prevalence 0.2222 0.039931 0.7002 0.03762
Detection Rate 0.1921 0.007523 0.6939 0.02373
Detection Prevalence 0.2367 0.013310 0.7228 0.02720
Balanced Accuracy 0.9036 0.591189 0.9472 0.81358
10 Naive Bayes 朴素贝叶斯
library(e1071)
#Build the model
model10<-naiveBayes(class~buying+maint+doors+persons+lug_boot+safety,data=car_eval,k=5)
#Summarize the model
summary(model10)
#Predict using the model
car_eval$pred_naive<-predict(model10,x)
#Accuracy of the model
mtab<-table(car_evalclass)
confusionMatrix(mtab)
Confusion Matrix and Statistics
acc good unacc vgood
acc 289 46 47 26
good 10 21 2 0
unacc 85 0 1161 0
vgood 0 2 0 39
Overall Statistics
Accuracy : 0.8738
95% CI : (0.8573, 0.8891)
No Information Rate : 0.7002
P-Value [Acc > NIR] : < 2.2e-16
Kappa : 0.7139
Mcnemar's Test P-Value : NA
Statistics by Class:
Class: acc Class: good Class: unacc Class: vgood
Sensitivity 0.7526 0.30435 0.9595 0.60000
Specificity 0.9115 0.99277 0.8359 0.99880
Pos Pred Value 0.7083 0.63636 0.9318 0.95122
Neg Pred Value 0.9280 0.97168 0.8983 0.98459
Prevalence 0.2222 0.03993 0.7002 0.03762
Detection Rate 0.1672 0.01215 0.6719 0.02257
Detection Prevalence 0.2361 0.01910 0.7211 0.02373
Balanced Accuracy 0.8320 0.64856 0.8977 0.79940
Non-Linear Classification with Decision Trees 非线性树模型
11 Classification and Regression Trees(CART) 分类回归树
library(rpart)
#Build the model
model11<-rpart(class~buying+maint+doors+persons+lug_boot+safety,data=car_eval)
#Summarize the model
summary(model11)
#Predict using the model
car_eval$pred_cart<-predict(model11,x,type="class")
#Accuracy of the model
mtab<-table(car_evalclass)
confusionMatrix(mtab)
Confusion Matrix and Statistics
acc good unacc vgood
acc 358 0 45 13
good 16 60 4 0
unacc 7 0 1161 0
vgood 3 9 0 52
Overall Statistics
Accuracy : 0.9439
95% CI : (0.9319, 0.9542)
No Information Rate : 0.7002
P-Value [Acc > NIR] : < 2.2e-16
Kappa : 0.8806
Mcnemar's Test P-Value : NA
Statistics by Class:
Class: acc Class: good Class: unacc Class: vgood
Sensitivity 0.9323 0.86957 0.9595 0.80000
Specificity 0.9568 0.98794 0.9865 0.99278
Pos Pred Value 0.8606 0.75000 0.9940 0.81250
Neg Pred Value 0.9802 0.99454 0.9125 0.99219
Prevalence 0.2222 0.03993 0.7002 0.03762
Detection Rate 0.2072 0.03472 0.6719 0.03009
Detection Prevalence 0.2407 0.04630 0.6759 0.03704
Balanced Accuracy 0.9446 0.92875 0.9730 0.89639
12 C4.5 C4.5树
library(caret)
library(RWeka)
#Build the model
model12<-J48(class~buying+maint+doors+persons+lug_boot+safety,data=car_eval)
#Summarize the model
summary(model12)
#Predict using the model
car_eval$pred_c45<-predict(model12,x)
#Accuracy of the model
mtab<-table(car_evalclass)
confusionMatrix(mtab)
Confusion Matrix and Statistics
acc good unacc vgood
acc 380 9 35 4
good 2 57 3 6
unacc 0 0 1172 0
vgood 2 3 0 55
Overall Statistics
Accuracy : 0.963
95% CI : (0.9529, 0.9714)
No Information Rate : 0.7002
P-Value [Acc > NIR] : < 2.2e-16
Kappa : 0.9207
Mcnemar's Test P-Value : NA
Statistics by Class:
Class: acc Class: good Class: unacc Class: vgood
Sensitivity 0.9896 0.82609 0.9686 0.84615
Specificity 0.9643 0.99337 1.0000 0.99699
Pos Pred Value 0.8879 0.83824 1.0000 0.91667
Neg Pred Value 0.9969 0.99277 0.9317 0.99400
Prevalence 0.2222 0.03993 0.7002 0.03762
Detection Rate 0.2199 0.03299 0.6782 0.03183
Detection Prevalence 0.2477 0.03935 0.6782 0.03472
Balanced Accuracy 0.9769 0.90973 0.9843 0.92157
13 PART 分区树
library(RWeka)
#Build the model
model13<-PART(class~buying+maint+doors+persons+lug_boot+safety,data=car_eval)
#Summarize the model
summary(model13)
#Predict using the model
car_eval$pred_part<-predict(model13,x)
#Accuracy of the model
mtab<-table(car_evalclass)
confusionMatrix(mtab)
Confusion Matrix and Statistics
acc good unacc vgood
acc 374 3 13 0
good 1 63 1 0
unacc 6 1 1196 0
vgood 3 2 0 65
Overall Statistics
Accuracy : 0.9826
95% CI : (0.9753, 0.9883)
No Information Rate : 0.7002
P-Value [Acc > NIR] : < 2.2e-16
Kappa : 0.9622
Mcnemar's Test P-Value : NA
Statistics by Class:
Class: acc Class: good Class: unacc Class: vgood
Sensitivity 0.9740 0.91304 0.9884 1.00000
Specificity 0.9881 0.99879 0.9865 0.99699
Pos Pred Value 0.9590 0.96923 0.9942 0.92857
Neg Pred Value 0.9925 0.99639 0.9733 1.00000
Prevalence 0.2222 0.03993 0.7002 0.03762
Detection Rate 0.2164 0.03646 0.6921 0.03762
Detection Prevalence 0.2257 0.03762 0.6962 0.04051
Balanced Accuracy 0.9810 0.95592 0.9875 0.99850
14 Bagging CART 背包树
library(ipred)
#Build the model
model14<-bagging(class~buying+maint+doors+persons+lug_boot+safety,data=car_eval)
#Summarize the model
summary(model14)
#Predict using the model
car_eval$pred_bagging<-predict(model14,x)
#Accuracy of the model
mtab<-table(car_evalclass)
confusionMatrix(mtab)
Confusion Matrix and Statistics
acc good unacc vgood
acc 384 0 0 0
good 0 69 0 0
unacc 0 0 1210 0
vgood 0 0 0 65
Overall Statistics
Accuracy : 1
95% CI : (0.9979, 1)
No Information Rate : 0.7002
P-Value [Acc > NIR] : < 2.2e-16
Kappa : 1
Mcnemar's Test P-Value : NA
Statistics by Class:
Class: acc Class: good Class: unacc Class: vgood
Sensitivity 1.0000 1.00000 1.0000 1.00000
Specificity 1.0000 1.00000 1.0000 1.00000
Pos Pred Value 1.0000 1.00000 1.0000 1.00000
Neg Pred Value 1.0000 1.00000 1.0000 1.00000
Prevalence 0.2222 0.03993 0.7002 0.03762
Detection Rate 0.2222 0.03993 0.7002 0.03762
Detection Prevalence 0.2222 0.03993 0.7002 0.03762
Balanced Accuracy 1.0000 1.00000 1.0000 1.00000
15 Random Forest 随机森林
library(randomForest)
#Build the model
model15<-randomForest(class~buying+maint+doors+persons+lug_boot+safety,data=car_eval)
#Summarize the model
summary(model15)
#Predict using the model
car_eval$pred_randomforest<-predict(model15,x)
#Accuracy of the model
mtab<-table(car_evalclass)
confusionMatrix(mtab)
Confusion Matrix and Statistics
acc good unacc vgood
acc 384 0 1 0
good 0 69 0 0
unacc 0 0 1209 0
vgood 0 0 0 65
Overall Statistics
Accuracy : 0.9994
95% CI : (0.9968, 1)
No Information Rate : 0.7002
P-Value [Acc > NIR] : < 2.2e-16
Kappa : 0.9987
Mcnemar's Test P-Value : NA
Statistics by Class:
Class: acc Class: good Class: unacc Class: vgood
Sensitivity 1.0000 1.00000 0.9992 1.00000
Specificity 0.9993 1.00000 1.0000 1.00000
Pos Pred Value 0.9974 1.00000 1.0000 1.00000
Neg Pred Value 1.0000 1.00000 0.9981 1.00000
Prevalence 0.2222 0.03993 0.7002 0.03762
Detection Rate 0.2222 0.03993 0.699
