HW 3 - LDA, QDA, KNN, Logistic Regression
Manuel S. H. Valles
March 1, 2019
Question 10
library(ISLR)## Warning: package 'ISLR' was built under R version 3.5.2#Im going alias weekly as W. For profiling purposes#
W<-WeeklyWhat is Weekly
Weekly is S&P 500 stock index data with the first variable, Year, btheeing year the observation was recorded. The next 5 variables , Lag1:Lag5, represent the percentage return for the previous week, return for the 2 weeks previous, … ,for 5 weeks previous.
Variable no. 7 is Volume, meaning the average number of daily shares traded in billions. Today, the 8th column, is the percentage return for this week, and Direction, the final and 9th column, is a factor with levels Down and Up indicating whether the market had a positive or negative return on a given week
A. Lets look at it
#Look at the structure#
str(W)## 'data.frame': 1089 obs. of 9 variables:
## $ Year : num 1990 1990 1990 1990 1990 1990 1990 1990 1990 1990 ...
## $ Lag1 : num 0.816 -0.27 -2.576 3.514 0.712 ...
## $ Lag2 : num 1.572 0.816 -0.27 -2.576 3.514 ...
## $ Lag3 : num -3.936 1.572 0.816 -0.27 -2.576 ...
## $ Lag4 : num -0.229 -3.936 1.572 0.816 -0.27 ...
## $ Lag5 : num -3.484 -0.229 -3.936 1.572 0.816 ...
## $ Volume : num 0.155 0.149 0.16 0.162 0.154 ...
## $ Today : num -0.27 -2.576 3.514 0.712 1.178 ...
## $ Direction: Factor w/ 2 levels "Down","Up": 1 1 2 2 2 1 2 2 2 1 ...#Peek at the dataset#
head(W)## Year Lag1 Lag2 Lag3 Lag4 Lag5 Volume Today Direction
## 1 1990 0.816 1.572 -3.936 -0.229 -3.484 0.1549760 -0.270 Down
## 2 1990 -0.270 0.816 1.572 -3.936 -0.229 0.1485740 -2.576 Down
## 3 1990 -2.576 -0.270 0.816 1.572 -3.936 0.1598375 3.514 Up
## 4 1990 3.514 -2.576 -0.270 0.816 1.572 0.1616300 0.712 Up
## 5 1990 0.712 3.514 -2.576 -0.270 0.816 0.1537280 1.178 Up
## 6 1990 1.178 0.712 3.514 -2.576 -0.270 0.1544440 -1.372 Down#Look at the correlations between numeric values#
corrplot::corrplot(cor(W[1:8]))
#Year and Volume stick out, lets look#
cor(W$Year,W$Volume)## [1] 0.8419416#It looks like they have a strong positive correlation#B. Logistic Regression
W.glm.fit<-glm(Direction~.-Year-Today, family = binomial, data = W)
#It looks like only Lag2 or return for the 2 weeks previous, is significant with a coefficient of .05844 and p values, 0.0296#
summary(W.glm.fit)##
## Call:
## glm(formula = Direction ~ . - Year - Today, family = binomial,
## data = W)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.6949 -1.2565 0.9913 1.0849 1.4579
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) 0.26686 0.08593 3.106 0.0019 **
## Lag1 -0.04127 0.02641 -1.563 0.1181
## Lag2 0.05844 0.02686 2.175 0.0296 *
## Lag3 -0.01606 0.02666 -0.602 0.5469
## Lag4 -0.02779 0.02646 -1.050 0.2937
## Lag5 -0.01447 0.02638 -0.549 0.5833
## Volume -0.02274 0.03690 -0.616 0.5377
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 1496.2 on 1088 degrees of freedom
## Residual deviance: 1486.4 on 1082 degrees of freedom
## AIC: 1500.4
##
## Number of Fisher Scoring iterations: 4C. Confusion Matrix for Logistic Regression
W.glm.fit<-glm(Direction~.-Year-Today, family = binomial, data = W)
W.probz <- predict(W.glm.fit,W, type = "response")
W.pred.glm <- rep("Down", length(W.probz))
W.pred.glm<-as.factor(ifelse(W.probz > 0.5,"Up","Down"))
caret::confusionMatrix(W.pred.glm, W$Direction)## Confusion Matrix and Statistics
##
## Reference
## Prediction Down Up
## Down 54 48
## Up 430 557
##
## Accuracy : 0.5611
## 95% CI : (0.531, 0.5908)
## No Information Rate : 0.5556
## P-Value [Acc > NIR] : 0.369
##
## Kappa : 0.035
## Mcnemar's Test P-Value : <2e-16
##
## Sensitivity : 0.11157
## Specificity : 0.92066
## Pos Pred Value : 0.52941
## Neg Pred Value : 0.56434
## Prevalence : 0.44444
## Detection Rate : 0.04959
## Detection Prevalence : 0.09366
## Balanced Accuracy : 0.51612
##
## 'Positive' Class : Down
## Model was accurate 56.11% of the time, which is slightly, but not significantly, higher than the 55.56% Null Information Rate. The model correctly predicted Down only 11.15% of the time, but successfully predicted Up 92.06% of the time
D.Log with Train and Test
#Here, I make the training set with the years between 1990 and 2008 and only keep the variables I need
train<-W[W$Year<=2008,c("Lag2","Direction")]
#And here i make the testing set with years 2009 and 2010.
test<-W[W$Year>2008,c("Lag2","Direction")]
#Model
W.log.train<-glm(Direction~., data = train, family = binomial)
#Confusion Matrix
testpredz<-predict(W.log.train, test,type='response')
tp_dirk<-rep('Down', length(testpredz))
tp_dirk<-as.factor(ifelse(testpredz > 0.5,"Up","Down"))
caret::confusionMatrix(tp_dirk, test$Direction)## Confusion Matrix and Statistics
##
## Reference
## Prediction Down Up
## Down 9 5
## Up 34 56
##
## Accuracy : 0.625
## 95% CI : (0.5247, 0.718)
## No Information Rate : 0.5865
## P-Value [Acc > NIR] : 0.2439
##
## Kappa : 0.1414
## Mcnemar's Test P-Value : 7.34e-06
##
## Sensitivity : 0.20930
## Specificity : 0.91803
## Pos Pred Value : 0.64286
## Neg Pred Value : 0.62222
## Prevalence : 0.41346
## Detection Rate : 0.08654
## Detection Prevalence : 0.13462
## Balanced Accuracy : 0.56367
##
## 'Positive' Class : Down
## E. Linear Discriminant Analysis
library(MASS)
W.lda.train<-lda(Direction~., data = train)
testpredz<-predict(W.lda.train, test,type='response')
tp_dirk<-testpredz$class
caret::confusionMatrix(table(tp_dirk, Test.Direction = test$Direction))## Confusion Matrix and Statistics
##
## Test.Direction
## tp_dirk Down Up
## Down 9 5
## Up 34 56
##
## Accuracy : 0.625
## 95% CI : (0.5247, 0.718)
## No Information Rate : 0.5865
## P-Value [Acc > NIR] : 0.2439
##
## Kappa : 0.1414
## Mcnemar's Test P-Value : 7.34e-06
##
## Sensitivity : 0.20930
## Specificity : 0.91803
## Pos Pred Value : 0.64286
## Neg Pred Value : 0.62222
## Prevalence : 0.41346
## Detection Rate : 0.08654
## Detection Prevalence : 0.13462
## Balanced Accuracy : 0.56367
##
## 'Positive' Class : Down
## F. Quadratic Discriminant Analysis
W.qda.train<-qda(Direction~., data = train)
testpredz<-predict(W.qda.train, test,type='response')
tp_dirk<-testpredz$class
caret::confusionMatrix(table(tp_dirk, Test.Direction = test$Direction))## Confusion Matrix and Statistics
##
## Test.Direction
## tp_dirk Down Up
## Down 0 0
## Up 43 61
##
## Accuracy : 0.5865
## 95% CI : (0.4858, 0.6823)
## No Information Rate : 0.5865
## P-Value [Acc > NIR] : 0.5419
##
## Kappa : 0
## Mcnemar's Test P-Value : 1.504e-10
##
## Sensitivity : 0.0000
## Specificity : 1.0000
## Pos Pred Value : NaN
## Neg Pred Value : 0.5865
## Prevalence : 0.4135
## Detection Rate : 0.0000
## Detection Prevalence : 0.0000
## Balanced Accuracy : 0.5000
##
## 'Positive' Class : Down
## G. KNN with K = 1
library(class)
#Get the x variable for train, and y variable for train
train.x<-as.matrix(W[W$Year<=2008,]$Lag2)
train.y<-W[W$Year<=2008,]$Direction
#Get the x variable for test, and y variable for test
test.x<-as.matrix(W[W$Year>2008,]$Lag2)
test.y<-W[W$Year>2008,]$Direction
knn.pred<-knn(train.x,test.x,train.y, k=1)
caret::confusionMatrix(knn.pred,test.y)## Confusion Matrix and Statistics
##
## Reference
## Prediction Down Up
## Down 21 29
## Up 22 32
##
## Accuracy : 0.5096
## 95% CI : (0.4097, 0.609)
## No Information Rate : 0.5865
## P-Value [Acc > NIR] : 0.9540
##
## Kappa : 0.0127
## Mcnemar's Test P-Value : 0.4008
##
## Sensitivity : 0.4884
## Specificity : 0.5246
## Pos Pred Value : 0.4200
## Neg Pred Value : 0.5926
## Prevalence : 0.4135
## Detection Rate : 0.2019
## Detection Prevalence : 0.4808
## Balanced Accuracy : 0.5065
##
## 'Positive' Class : Down
## H. Which model worked best?
The worst model may have been KNN, with an accuracy of 50.96%, but QDA is just disappointing. We’ve come too far as human beings for QDA to just pick “Up” on all the damn factors. It was more accurate that KNN, with 58.65% accuracy, but to see a result like that… .Both LDA and Logistical Regression had an accuracy of 62.5%, and got over 90% of the “Up”s right. Despite this, the model was not significantly more accurate than using the Null Information rate
Question 11
A. Create binary variable, mpg10
a<-Auto
a$mpg01<-ifelse(a$mpg>median(a$mpg),1,0)B. Explore Auto
Auto is a dataset with data on 392 cars. They documented 9 variables per car: mpg, cylinders, displacement, horsepower, weight, acceleration, year, origin, and name
cor(a[,-9],a$mpg01)## [,1]
## mpg 0.8369392
## cylinders -0.7591939
## displacement -0.7534766
## horsepower -0.6670526
## weight -0.7577566
## acceleration 0.3468215
## year 0.4299042
## origin 0.5136984
## mpg01 1.0000000corrplot::corrplot(cor(a[,-9],a$mpg01))
Cylinders, displacement, horsepower, and weight all have strong negative correlations
Acceleration, year , and origin have positive correlations, but they are only .347,.430, and .514
library(car)## Warning: package 'car' was built under R version 3.5.2## Loading required package: carData## Warning: package 'carData' was built under R version 3.5.2library(lattice)
super.sym <- trellis.par.get("superpose.symbol")
splom(a[c(-1,-9,-10)], groups=a$mpg01, data=a, panel= panel.superpose,key=list(title="Over/Under MPG Median", columns=2,points=list(pch=super.sym$pch[1:2], col=super.sym$col[1:2]) ,text=list(c('Over MPG Median','Under MPG Median'))))## Warning in splom.data.frame(a[c(-1, -9, -10)], groups = a$mpg01, data =
## a, : explicit 'data' specification ignored
C. Splitting Data
a$origin<-as.factor(a$origin)
a$mpg01<-as.factor(a$mpg01)
set.seed(2020)
train<-sample(nrow(a),nrow(a)*.8)
a.train<-a[train,c(-1,-9)]
a.test<-a[-train,c(-1,-9)]D. LDA
#Im going to use the variables: cylinders, displacement, horsepower, weight, acceleration, year, and origin as the independent variables. mpg01 is the dependent variable
a.lda.train<-lda(mpg01~., data = a.train)
testpredz<-predict(a.lda.train, a.test,type='response')
tp_mpg<-testpredz$class
caret::confusionMatrix(tp_mpg, a.test$mpg01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 33 3
## 1 6 37
##
## Accuracy : 0.8861
## 95% CI : (0.7947, 0.9466)
## No Information Rate : 0.5063
## P-Value [Acc > NIR] : 8.396e-13
##
## Kappa : 0.7719
## Mcnemar's Test P-Value : 0.505
##
## Sensitivity : 0.8462
## Specificity : 0.9250
## Pos Pred Value : 0.9167
## Neg Pred Value : 0.8605
## Prevalence : 0.4937
## Detection Rate : 0.4177
## Detection Prevalence : 0.4557
## Balanced Accuracy : 0.8856
##
## 'Positive' Class : 0
## Accuracy : 0.8861
E. QDA
a.qda.train<-qda(mpg01~., data = a.train)
confit<-function(model,test_ds, test_ds_testvar){
testpredz<-predict(model, test_ds,type='response')
caret::confusionMatrix(testpredz$class, test_ds_testvar)
}
confit(a.qda.train,a.test,a.test$mpg01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 35 3
## 1 4 37
##
## Accuracy : 0.9114
## 95% CI : (0.8259, 0.9636)
## No Information Rate : 0.5063
## P-Value [Acc > NIR] : 1.201e-14
##
## Kappa : 0.8227
## Mcnemar's Test P-Value : 1
##
## Sensitivity : 0.8974
## Specificity : 0.9250
## Pos Pred Value : 0.9211
## Neg Pred Value : 0.9024
## Prevalence : 0.4937
## Detection Rate : 0.4430
## Detection Prevalence : 0.4810
## Balanced Accuracy : 0.9112
##
## 'Positive' Class : 0
## Accuracy : 0.9114
F. Logistical Regression
a.log.train<-glm(mpg01~., data = a.train, family = 'binomial')
testpredz<-predict(a.log.train, a.test,type='response')
a.pred.glm<-as.factor(ifelse(testpredz>.5,1,0))
caret::confusionMatrix(a.pred.glm, a.test$mpg01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 35 6
## 1 4 34
##
## Accuracy : 0.8734
## 95% CI : (0.7795, 0.9376)
## No Information Rate : 0.5063
## P-Value [Acc > NIR] : 5.838e-12
##
## Kappa : 0.747
## Mcnemar's Test P-Value : 0.7518
##
## Sensitivity : 0.8974
## Specificity : 0.8500
## Pos Pred Value : 0.8537
## Neg Pred Value : 0.8947
## Prevalence : 0.4937
## Detection Rate : 0.4430
## Detection Prevalence : 0.5190
## Balanced Accuracy : 0.8737
##
## 'Positive' Class : 0
## Accuracy : 0.8734
G. KNN k=1
train.y<-(a.train[,8])
train.x<-(a.train[,-8])
test.x<-(a.test[,-8])
test.y<-(a.test[,8])
knn.pred<-class::knn(train.x,test.x,train.y, k=1)
caret::confusionMatrix(knn.pred,a.test$mpg01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 33 7
## 1 6 33
##
## Accuracy : 0.8354
## 95% CI : (0.7351, 0.9094)
## No Information Rate : 0.5063
## P-Value [Acc > NIR] : 1.054e-09
##
## Kappa : 0.6709
## Mcnemar's Test P-Value : 1
##
## Sensitivity : 0.8462
## Specificity : 0.8250
## Pos Pred Value : 0.8250
## Neg Pred Value : 0.8462
## Prevalence : 0.4937
## Detection Rate : 0.4177
## Detection Prevalence : 0.5063
## Balanced Accuracy : 0.8356
##
## 'Positive' Class : 0
## Accuracy : 0.8354
G. k=2
knn.pred<-class::knn(train.x,test.x,train.y, k=2)
caret::confusionMatrix(knn.pred,a.test$mpg01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 33 9
## 1 6 31
##
## Accuracy : 0.8101
## 95% CI : (0.7062, 0.8897)
## No Information Rate : 0.5063
## P-Value [Acc > NIR] : 2.144e-08
##
## Kappa : 0.6206
## Mcnemar's Test P-Value : 0.6056
##
## Sensitivity : 0.8462
## Specificity : 0.7750
## Pos Pred Value : 0.7857
## Neg Pred Value : 0.8378
## Prevalence : 0.4937
## Detection Rate : 0.4177
## Detection Prevalence : 0.5316
## Balanced Accuracy : 0.8106
##
## 'Positive' Class : 0
## Accuracy : 0.8101
G. k=5
knn.pred<-class::knn(train.x,test.x,train.y, k=5)
caret::confusionMatrix(knn.pred,a.test$mpg01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 34 8
## 1 5 32
##
## Accuracy : 0.8354
## 95% CI : (0.7351, 0.9094)
## No Information Rate : 0.5063
## P-Value [Acc > NIR] : 1.054e-09
##
## Kappa : 0.6711
## Mcnemar's Test P-Value : 0.5791
##
## Sensitivity : 0.8718
## Specificity : 0.8000
## Pos Pred Value : 0.8095
## Neg Pred Value : 0.8649
## Prevalence : 0.4937
## Detection Rate : 0.4304
## Detection Prevalence : 0.5316
## Balanced Accuracy : 0.8359
##
## 'Positive' Class : 0
## Accuracy : 0.8354
G. k=10
knn.pred<-class::knn(train.x,test.x,train.y, k=1)
caret::confusionMatrix(knn.pred,a.test$mpg01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 33 7
## 1 6 33
##
## Accuracy : 0.8354
## 95% CI : (0.7351, 0.9094)
## No Information Rate : 0.5063
## P-Value [Acc > NIR] : 1.054e-09
##
## Kappa : 0.6709
## Mcnemar's Test P-Value : 1
##
## Sensitivity : 0.8462
## Specificity : 0.8250
## Pos Pred Value : 0.8250
## Neg Pred Value : 0.8462
## Prevalence : 0.4937
## Detection Rate : 0.4177
## Detection Prevalence : 0.5063
## Balanced Accuracy : 0.8356
##
## 'Positive' Class : 0
## Accuracy : 0.8354
Question 13
Boston
Boston contains housing data in Boston. It has the following columns:
crim-per capita crime rate by town. zn-proportion of residential land zoned for lots over 25,000 sq.ft. indus-proportion of non-retail business acres per town. chas-Charles River dummy variable (= 1 if tract bounds river; 0 otherwise). nox-nitrogen oxides concentration (parts per 10 million). rm-average number of rooms per dwelling. age-proportion of owner-occupied units built prior to 1940. dis-weighted mean of distances to five Boston employment centres. rad-index of accessibility to radial highways. tax-full-value property-tax rate per $10,000. ptratio-pupil-teacher ratio by town. black-1000(Bk - 0.63)^2 where Bk is the proportion of blacks by town. lstat-lower status of the population (percent). medv-median value of owner-occupied homes in $1000s
Prep and Peek at Boston
b<-Boston
b$crim01<-ifelse(b$crim>median(b$crim),1,0)
cor(b[,-15],b$crim01)## [,1]
## crim 0.40939545
## zn -0.43615103
## indus 0.60326017
## chas 0.07009677
## nox 0.72323480
## rm -0.15637178
## age 0.61393992
## dis -0.61634164
## rad 0.61978625
## tax 0.60874128
## ptratio 0.25356836
## black -0.35121093
## lstat 0.45326273
## medv -0.26301673Selecting our variables
b$crim01<-as.factor(ifelse(b$crim>median(b$crim),1,0))
#Goes from all, to those with a correlation greater than .4, and then to those with a correlation > .6.#
var<-1 #do -var #
var1<-c('zn','indus','nox','age','dis','rad','tax','lstat','crim01')
var2<-c('indus','nox','age','dis','rad','tax','crim01')
#Making samples#
set.seed(111)
train<-sample(nrow(b),nrow(b)*.8)
train1<-sample(nrow(b),nrow(b)*.7)
train2<-sample(nrow(b),nrow(b)*.825)Making our various sets of test and train
Here we aim to try our different models, with multiple sets of data.
#First set of Var#
b.train<-b[train,-var]
b.test<-b[-train,-var]
b.train01<-b[train1,-var]
b.test01<-b[-train1,-var]
b.train02<-b[train2,-var]
b.test02<-b[-train2,-var]
#Second set of Var#
b.train1<-b[train1,var1]
b.test1<-b[-train1,var1]
b.train10<-b[train,var1]
b.test10<-b[-train,var1]
b.train12<-b[train2,var1]
b.test12<-b[-train2,var1]
#Third Set of Variables
b.train2<-b[train2,var2]
b.test2<-b[-train2,var2]
b.train20<-b[train,var2]
b.test20<-b[-train,var2]
b.train21<-b[train1,var2]
b.test21<-b[-train1,var2]LDA
b.lda.train<-lda(crim01 ~ ., data = b.train)
confit(b.lda.train,b.test,b.test$crim01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 46 13
## 1 1 42
##
## Accuracy : 0.8627
## 95% CI : (0.7804, 0.9229)
## No Information Rate : 0.5392
## P-Value [Acc > NIR] : 3.507e-12
##
## Kappa : 0.7288
## Mcnemar's Test P-Value : 0.003283
##
## Sensitivity : 0.9787
## Specificity : 0.7636
## Pos Pred Value : 0.7797
## Neg Pred Value : 0.9767
## Prevalence : 0.4608
## Detection Rate : 0.4510
## Detection Prevalence : 0.5784
## Balanced Accuracy : 0.8712
##
## 'Positive' Class : 0
## b.lda.train01<-lda(crim01 ~ ., data = b.train01)
confit(b.lda.train01,b.test01,b.test01$crim01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 70 20
## 1 2 60
##
## Accuracy : 0.8553
## 95% CI : (0.7891, 0.907)
## No Information Rate : 0.5263
## P-Value [Acc > NIR] : < 2.2e-16
##
## Kappa : 0.7133
## Mcnemar's Test P-Value : 0.0002896
##
## Sensitivity : 0.9722
## Specificity : 0.7500
## Pos Pred Value : 0.7778
## Neg Pred Value : 0.9677
## Prevalence : 0.4737
## Detection Rate : 0.4605
## Detection Prevalence : 0.5921
## Balanced Accuracy : 0.8611
##
## 'Positive' Class : 0
## b.lda.train02<-lda(crim01 ~ ., data = b.train02)
confit(b.lda.train02,b.test02,b.test02$crim01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 33 9
## 1 5 42
##
## Accuracy : 0.8427
## 95% CI : (0.7502, 0.9112)
## No Information Rate : 0.573
## P-Value [Acc > NIR] : 4.914e-08
##
## Kappa : 0.6828
## Mcnemar's Test P-Value : 0.4227
##
## Sensitivity : 0.8684
## Specificity : 0.8235
## Pos Pred Value : 0.7857
## Neg Pred Value : 0.8936
## Prevalence : 0.4270
## Detection Rate : 0.3708
## Detection Prevalence : 0.4719
## Balanced Accuracy : 0.8460
##
## 'Positive' Class : 0
## b.lda.train1<-lda(crim01 ~ ., data = b.train1)
confit(b.lda.train1,b.test1,b.test1$crim01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 68 20
## 1 4 60
##
## Accuracy : 0.8421
## 95% CI : (0.7742, 0.8961)
## No Information Rate : 0.5263
## P-Value [Acc > NIR] : 2.34e-16
##
## Kappa : 0.6868
## Mcnemar's Test P-Value : 0.0022
##
## Sensitivity : 0.9444
## Specificity : 0.7500
## Pos Pred Value : 0.7727
## Neg Pred Value : 0.9375
## Prevalence : 0.4737
## Detection Rate : 0.4474
## Detection Prevalence : 0.5789
## Balanced Accuracy : 0.8472
##
## 'Positive' Class : 0
## b.lda.train10<-lda(crim01 ~ ., data = b.train10)
confit(b.lda.train10,b.test10,b.test10$crim01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 46 13
## 1 1 42
##
## Accuracy : 0.8627
## 95% CI : (0.7804, 0.9229)
## No Information Rate : 0.5392
## P-Value [Acc > NIR] : 3.507e-12
##
## Kappa : 0.7288
## Mcnemar's Test P-Value : 0.003283
##
## Sensitivity : 0.9787
## Specificity : 0.7636
## Pos Pred Value : 0.7797
## Neg Pred Value : 0.9767
## Prevalence : 0.4608
## Detection Rate : 0.4510
## Detection Prevalence : 0.5784
## Balanced Accuracy : 0.8712
##
## 'Positive' Class : 0
## b.lda.train12<-lda(crim01 ~ ., data = b.train12)
confit(b.lda.train12,b.test12,b.test12$crim01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 31 12
## 1 7 39
##
## Accuracy : 0.7865
## 95% CI : (0.6869, 0.8663)
## No Information Rate : 0.573
## P-Value [Acc > NIR] : 1.948e-05
##
## Kappa : 0.5709
## Mcnemar's Test P-Value : 0.3588
##
## Sensitivity : 0.8158
## Specificity : 0.7647
## Pos Pred Value : 0.7209
## Neg Pred Value : 0.8478
## Prevalence : 0.4270
## Detection Rate : 0.3483
## Detection Prevalence : 0.4831
## Balanced Accuracy : 0.7902
##
## 'Positive' Class : 0
## b.lda.train2<-lda(crim01 ~ ., data = b.train2)
confit(b.lda.train2,b.test2,b.test2$crim01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 32 12
## 1 6 39
##
## Accuracy : 0.7978
## 95% CI : (0.6993, 0.8755)
## No Information Rate : 0.573
## P-Value [Acc > NIR] : 6.765e-06
##
## Kappa : 0.5948
## Mcnemar's Test P-Value : 0.2386
##
## Sensitivity : 0.8421
## Specificity : 0.7647
## Pos Pred Value : 0.7273
## Neg Pred Value : 0.8667
## Prevalence : 0.4270
## Detection Rate : 0.3596
## Detection Prevalence : 0.4944
## Balanced Accuracy : 0.8034
##
## 'Positive' Class : 0
## b.lda.train20<-lda(crim01 ~ ., data = b.train20)
confit(b.lda.train20,b.test20,b.test20$crim01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 45 12
## 1 2 43
##
## Accuracy : 0.8627
## 95% CI : (0.7804, 0.9229)
## No Information Rate : 0.5392
## P-Value [Acc > NIR] : 3.507e-12
##
## Kappa : 0.728
## Mcnemar's Test P-Value : 0.01616
##
## Sensitivity : 0.9574
## Specificity : 0.7818
## Pos Pred Value : 0.7895
## Neg Pred Value : 0.9556
## Prevalence : 0.4608
## Detection Rate : 0.4412
## Detection Prevalence : 0.5588
## Balanced Accuracy : 0.8696
##
## 'Positive' Class : 0
## b.lda.train21<-lda(crim01 ~ ., data = b.train21)
confit(b.lda.train21,b.test21,b.test21$crim01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 69 20
## 1 3 60
##
## Accuracy : 0.8487
## 95% CI : (0.7817, 0.9016)
## No Information Rate : 0.5263
## P-Value [Acc > NIR] : < 2.2e-16
##
## Kappa : 0.7001
## Mcnemar's Test P-Value : 0.0008492
##
## Sensitivity : 0.9583
## Specificity : 0.7500
## Pos Pred Value : 0.7753
## Neg Pred Value : 0.9524
## Prevalence : 0.4737
## Detection Rate : 0.4539
## Detection Prevalence : 0.5855
## Balanced Accuracy : 0.8542
##
## 'Positive' Class : 0
## QDA
b.qda.train<-qda(crim01 ~ ., data = b.train)
confit(b.qda.train,b.test,b.test$crim01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 46 11
## 1 1 44
##
## Accuracy : 0.8824
## 95% CI : (0.8035, 0.9377)
## No Information Rate : 0.5392
## P-Value [Acc > NIR] : 1.054e-13
##
## Kappa : 0.7669
## Mcnemar's Test P-Value : 0.009375
##
## Sensitivity : 0.9787
## Specificity : 0.8000
## Pos Pred Value : 0.8070
## Neg Pred Value : 0.9778
## Prevalence : 0.4608
## Detection Rate : 0.4510
## Detection Prevalence : 0.5588
## Balanced Accuracy : 0.8894
##
## 'Positive' Class : 0
## b.qda.train01<-qda(crim01 ~ ., data = b.train01)
confit(b.qda.train01,b.test01,b.test01$crim01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 71 14
## 1 1 66
##
## Accuracy : 0.9013
## 95% CI : (0.8425, 0.9437)
## No Information Rate : 0.5263
## P-Value [Acc > NIR] : < 2.2e-16
##
## Kappa : 0.8039
## Mcnemar's Test P-Value : 0.001946
##
## Sensitivity : 0.9861
## Specificity : 0.8250
## Pos Pred Value : 0.8353
## Neg Pred Value : 0.9851
## Prevalence : 0.4737
## Detection Rate : 0.4671
## Detection Prevalence : 0.5592
## Balanced Accuracy : 0.9056
##
## 'Positive' Class : 0
## b.qda.train02<-qda(crim01 ~ ., data = b.train02)
confit(b.qda.train02,b.test02,b.test02$crim01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 37 8
## 1 1 43
##
## Accuracy : 0.8989
## 95% CI : (0.8167, 0.9527)
## No Information Rate : 0.573
## P-Value [Acc > NIR] : 1.584e-11
##
## Kappa : 0.7981
## Mcnemar's Test P-Value : 0.0455
##
## Sensitivity : 0.9737
## Specificity : 0.8431
## Pos Pred Value : 0.8222
## Neg Pred Value : 0.9773
## Prevalence : 0.4270
## Detection Rate : 0.4157
## Detection Prevalence : 0.5056
## Balanced Accuracy : 0.9084
##
## 'Positive' Class : 0
## b.qda.train1<-qda(crim01 ~ ., data = b.train1)
confit(b.qda.train1,b.test1,b.test1$crim01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 69 11
## 1 3 69
##
## Accuracy : 0.9079
## 95% CI : (0.8503, 0.9487)
## No Information Rate : 0.5263
## P-Value [Acc > NIR] : < 2e-16
##
## Kappa : 0.8163
## Mcnemar's Test P-Value : 0.06137
##
## Sensitivity : 0.9583
## Specificity : 0.8625
## Pos Pred Value : 0.8625
## Neg Pred Value : 0.9583
## Prevalence : 0.4737
## Detection Rate : 0.4539
## Detection Prevalence : 0.5263
## Balanced Accuracy : 0.9104
##
## 'Positive' Class : 0
## b.qda.train10<-qda(crim01 ~ ., data = b.train10)
confit(b.qda.train10,b.test10,b.test10$crim01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 47 6
## 1 0 49
##
## Accuracy : 0.9412
## 95% CI : (0.8764, 0.9781)
## No Information Rate : 0.5392
## P-Value [Acc > NIR] : < 2e-16
##
## Kappa : 0.8827
## Mcnemar's Test P-Value : 0.04123
##
## Sensitivity : 1.0000
## Specificity : 0.8909
## Pos Pred Value : 0.8868
## Neg Pred Value : 1.0000
## Prevalence : 0.4608
## Detection Rate : 0.4608
## Detection Prevalence : 0.5196
## Balanced Accuracy : 0.9455
##
## 'Positive' Class : 0
## b.qda.train12<-qda(crim01 ~ ., data = b.train12)
confit(b.qda.train12,b.test12,b.test12$crim01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 35 7
## 1 3 44
##
## Accuracy : 0.8876
## 95% CI : (0.8031, 0.9448)
## No Information Rate : 0.573
## P-Value [Acc > NIR] : 9.646e-11
##
## Kappa : 0.7734
## Mcnemar's Test P-Value : 0.3428
##
## Sensitivity : 0.9211
## Specificity : 0.8627
## Pos Pred Value : 0.8333
## Neg Pred Value : 0.9362
## Prevalence : 0.4270
## Detection Rate : 0.3933
## Detection Prevalence : 0.4719
## Balanced Accuracy : 0.8919
##
## 'Positive' Class : 0
## b.qda.train2<-qda(crim01 ~ ., data = b.train2)
confit(b.qda.train2,b.test2,b.test2$crim01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 35 7
## 1 3 44
##
## Accuracy : 0.8876
## 95% CI : (0.8031, 0.9448)
## No Information Rate : 0.573
## P-Value [Acc > NIR] : 9.646e-11
##
## Kappa : 0.7734
## Mcnemar's Test P-Value : 0.3428
##
## Sensitivity : 0.9211
## Specificity : 0.8627
## Pos Pred Value : 0.8333
## Neg Pred Value : 0.9362
## Prevalence : 0.4270
## Detection Rate : 0.3933
## Detection Prevalence : 0.4719
## Balanced Accuracy : 0.8919
##
## 'Positive' Class : 0
## b.qda.train20<-qda(crim01 ~ ., data = b.train20)
confit(b.qda.train20,b.test20,b.test20$crim01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 47 9
## 1 0 46
##
## Accuracy : 0.9118
## 95% CI : (0.8391, 0.9589)
## No Information Rate : 0.5392
## P-Value [Acc > NIR] : 2.73e-16
##
## Kappa : 0.8249
## Mcnemar's Test P-Value : 0.007661
##
## Sensitivity : 1.0000
## Specificity : 0.8364
## Pos Pred Value : 0.8393
## Neg Pred Value : 1.0000
## Prevalence : 0.4608
## Detection Rate : 0.4608
## Detection Prevalence : 0.5490
## Balanced Accuracy : 0.9182
##
## 'Positive' Class : 0
## b.qda.train21<-qda(crim01 ~ ., data = b.train21)
confit(b.qda.train21,b.test21,b.test21$crim01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 71 14
## 1 1 66
##
## Accuracy : 0.9013
## 95% CI : (0.8425, 0.9437)
## No Information Rate : 0.5263
## P-Value [Acc > NIR] : < 2.2e-16
##
## Kappa : 0.8039
## Mcnemar's Test P-Value : 0.001946
##
## Sensitivity : 0.9861
## Specificity : 0.8250
## Pos Pred Value : 0.8353
## Neg Pred Value : 0.9851
## Prevalence : 0.4737
## Detection Rate : 0.4671
## Detection Prevalence : 0.5592
## Balanced Accuracy : 0.9056
##
## 'Positive' Class : 0
## Logistic Regression
b.log.confit<-function(train, test){
b.log.train<-glm(crim01~., data = train, family = 'binomial')
testpredz<-predict(b.log.train, test,type='response')
b.pred.glm<-as.factor(ifelse(testpredz>.5,1,0))
caret::confusionMatrix(b.pred.glm, test$crim01)
}
b.log.confit(b.train,b.test)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 47 2
## 1 0 53
##
## Accuracy : 0.9804
## 95% CI : (0.931, 0.9976)
## No Information Rate : 0.5392
## P-Value [Acc > NIR] : <2e-16
##
## Kappa : 0.9607
## Mcnemar's Test P-Value : 0.4795
##
## Sensitivity : 1.0000
## Specificity : 0.9636
## Pos Pred Value : 0.9592
## Neg Pred Value : 1.0000
## Prevalence : 0.4608
## Detection Rate : 0.4608
## Detection Prevalence : 0.4804
## Balanced Accuracy : 0.9818
##
## 'Positive' Class : 0
## b.log.confit(b.train01,b.test01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 70 7
## 1 2 73
##
## Accuracy : 0.9408
## 95% CI : (0.8906, 0.9726)
## No Information Rate : 0.5263
## P-Value [Acc > NIR] : <2e-16
##
## Kappa : 0.8817
## Mcnemar's Test P-Value : 0.1824
##
## Sensitivity : 0.9722
## Specificity : 0.9125
## Pos Pred Value : 0.9091
## Neg Pred Value : 0.9733
## Prevalence : 0.4737
## Detection Rate : 0.4605
## Detection Prevalence : 0.5066
## Balanced Accuracy : 0.9424
##
## 'Positive' Class : 0
## b.log.confit(b.train02,b.test02)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 33 6
## 1 5 45
##
## Accuracy : 0.8764
## 95% CI : (0.7896, 0.9367)
## No Information Rate : 0.573
## P-Value [Acc > NIR] : 5.279e-10
##
## Kappa : 0.7483
## Mcnemar's Test P-Value : 1
##
## Sensitivity : 0.8684
## Specificity : 0.8824
## Pos Pred Value : 0.8462
## Neg Pred Value : 0.9000
## Prevalence : 0.4270
## Detection Rate : 0.3708
## Detection Prevalence : 0.4382
## Balanced Accuracy : 0.8754
##
## 'Positive' Class : 0
## b.log.confit(b.train1,b.test1)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 67 9
## 1 5 71
##
## Accuracy : 0.9079
## 95% CI : (0.8503, 0.9487)
## No Information Rate : 0.5263
## P-Value [Acc > NIR] : <2e-16
##
## Kappa : 0.8158
## Mcnemar's Test P-Value : 0.4227
##
## Sensitivity : 0.9306
## Specificity : 0.8875
## Pos Pred Value : 0.8816
## Neg Pred Value : 0.9342
## Prevalence : 0.4737
## Detection Rate : 0.4408
## Detection Prevalence : 0.5000
## Balanced Accuracy : 0.9090
##
## 'Positive' Class : 0
## b.log.confit(b.train10,b.test10)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 39 3
## 1 8 52
##
## Accuracy : 0.8922
## 95% CI : (0.8152, 0.9449)
## No Information Rate : 0.5392
## P-Value [Acc > NIR] : 1.6e-14
##
## Kappa : 0.7813
## Mcnemar's Test P-Value : 0.2278
##
## Sensitivity : 0.8298
## Specificity : 0.9455
## Pos Pred Value : 0.9286
## Neg Pred Value : 0.8667
## Prevalence : 0.4608
## Detection Rate : 0.3824
## Detection Prevalence : 0.4118
## Balanced Accuracy : 0.8876
##
## 'Positive' Class : 0
## b.log.confit(b.train12,b.test12)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 29 8
## 1 9 43
##
## Accuracy : 0.809
## 95% CI : (0.7119, 0.8846)
## No Information Rate : 0.573
## P-Value [Acc > NIR] : 2.199e-06
##
## Kappa : 0.6083
## Mcnemar's Test P-Value : 1
##
## Sensitivity : 0.7632
## Specificity : 0.8431
## Pos Pred Value : 0.7838
## Neg Pred Value : 0.8269
## Prevalence : 0.4270
## Detection Rate : 0.3258
## Detection Prevalence : 0.4157
## Balanced Accuracy : 0.8031
##
## 'Positive' Class : 0
## b.log.confit(b.train2,b.test2)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 31 6
## 1 7 45
##
## Accuracy : 0.8539
## 95% CI : (0.7632, 0.9199)
## No Information Rate : 0.573
## P-Value [Acc > NIR] : 1.184e-08
##
## Kappa : 0.7005
## Mcnemar's Test P-Value : 1
##
## Sensitivity : 0.8158
## Specificity : 0.8824
## Pos Pred Value : 0.8378
## Neg Pred Value : 0.8654
## Prevalence : 0.4270
## Detection Rate : 0.3483
## Detection Prevalence : 0.4157
## Balanced Accuracy : 0.8491
##
## 'Positive' Class : 0
## b.log.confit(b.train20,b.test20)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 40 5
## 1 7 50
##
## Accuracy : 0.8824
## 95% CI : (0.8035, 0.9377)
## No Information Rate : 0.5392
## P-Value [Acc > NIR] : 1.054e-13
##
## Kappa : 0.7625
## Mcnemar's Test P-Value : 0.7728
##
## Sensitivity : 0.8511
## Specificity : 0.9091
## Pos Pred Value : 0.8889
## Neg Pred Value : 0.8772
## Prevalence : 0.4608
## Detection Rate : 0.3922
## Detection Prevalence : 0.4412
## Balanced Accuracy : 0.8801
##
## 'Positive' Class : 0
## b.log.confit(b.train21,b.test21)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 68 14
## 1 4 66
##
## Accuracy : 0.8816
## 95% CI : (0.8193, 0.9283)
## No Information Rate : 0.5263
## P-Value [Acc > NIR] : < 2e-16
##
## Kappa : 0.7641
## Mcnemar's Test P-Value : 0.03389
##
## Sensitivity : 0.9444
## Specificity : 0.8250
## Pos Pred Value : 0.8293
## Neg Pred Value : 0.9429
## Prevalence : 0.4737
## Detection Rate : 0.4474
## Detection Prevalence : 0.5395
## Balanced Accuracy : 0.8847
##
## 'Positive' Class : 0
## KNN=1
b.knn.confit<-function(train,test){
train.y<-(train[,"crim01"])
train.x<-(train[,names(train)!="crim01"])
test.x<-(test[,names(test)!="crim01"])
test.y<-(test[,"crim01"])
knn.pred<-class::knn(train.x,test.x,train.y, k=1)
caret::confusionMatrix(knn.pred,test$crim01)
}
b.knn.confit(b.train,b.test)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 43 2
## 1 4 53
##
## Accuracy : 0.9412
## 95% CI : (0.8764, 0.9781)
## No Information Rate : 0.5392
## P-Value [Acc > NIR] : <2e-16
##
## Kappa : 0.8813
## Mcnemar's Test P-Value : 0.6831
##
## Sensitivity : 0.9149
## Specificity : 0.9636
## Pos Pred Value : 0.9556
## Neg Pred Value : 0.9298
## Prevalence : 0.4608
## Detection Rate : 0.4216
## Detection Prevalence : 0.4412
## Balanced Accuracy : 0.9393
##
## 'Positive' Class : 0
## b.knn.confit(b.train01,b.test01)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 70 6
## 1 2 74
##
## Accuracy : 0.9474
## 95% CI : (0.8989, 0.977)
## No Information Rate : 0.5263
## P-Value [Acc > NIR] : <2e-16
##
## Kappa : 0.8947
## Mcnemar's Test P-Value : 0.2888
##
## Sensitivity : 0.9722
## Specificity : 0.9250
## Pos Pred Value : 0.9211
## Neg Pred Value : 0.9737
## Prevalence : 0.4737
## Detection Rate : 0.4605
## Detection Prevalence : 0.5000
## Balanced Accuracy : 0.9486
##
## 'Positive' Class : 0
## b.knn.confit(b.train02,b.test02)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 32 6
## 1 6 45
##
## Accuracy : 0.8652
## 95% CI : (0.7763, 0.9283)
## No Information Rate : 0.573
## P-Value [Acc > NIR] : 2.617e-09
##
## Kappa : 0.7245
## Mcnemar's Test P-Value : 1
##
## Sensitivity : 0.8421
## Specificity : 0.8824
## Pos Pred Value : 0.8421
## Neg Pred Value : 0.8824
## Prevalence : 0.4270
## Detection Rate : 0.3596
## Detection Prevalence : 0.4270
## Balanced Accuracy : 0.8622
##
## 'Positive' Class : 0
## b.knn.confit(b.train1,b.test1)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 67 4
## 1 5 76
##
## Accuracy : 0.9408
## 95% CI : (0.8906, 0.9726)
## No Information Rate : 0.5263
## P-Value [Acc > NIR] : <2e-16
##
## Kappa : 0.8812
## Mcnemar's Test P-Value : 1
##
## Sensitivity : 0.9306
## Specificity : 0.9500
## Pos Pred Value : 0.9437
## Neg Pred Value : 0.9383
## Prevalence : 0.4737
## Detection Rate : 0.4408
## Detection Prevalence : 0.4671
## Balanced Accuracy : 0.9403
##
## 'Positive' Class : 0
## b.knn.confit(b.train10,b.test10)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 45 1
## 1 2 54
##
## Accuracy : 0.9706
## 95% CI : (0.9164, 0.9939)
## No Information Rate : 0.5392
## P-Value [Acc > NIR] : <2e-16
##
## Kappa : 0.9407
## Mcnemar's Test P-Value : 1
##
## Sensitivity : 0.9574
## Specificity : 0.9818
## Pos Pred Value : 0.9783
## Neg Pred Value : 0.9643
## Prevalence : 0.4608
## Detection Rate : 0.4412
## Detection Prevalence : 0.4510
## Balanced Accuracy : 0.9696
##
## 'Positive' Class : 0
## b.knn.confit(b.train12,b.test12)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 31 4
## 1 7 47
##
## Accuracy : 0.8764
## 95% CI : (0.7896, 0.9367)
## No Information Rate : 0.573
## P-Value [Acc > NIR] : 5.279e-10
##
## Kappa : 0.7449
## Mcnemar's Test P-Value : 0.5465
##
## Sensitivity : 0.8158
## Specificity : 0.9216
## Pos Pred Value : 0.8857
## Neg Pred Value : 0.8704
## Prevalence : 0.4270
## Detection Rate : 0.3483
## Detection Prevalence : 0.3933
## Balanced Accuracy : 0.8687
##
## 'Positive' Class : 0
## b.knn.confit(b.train2,b.test2)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 34 5
## 1 4 46
##
## Accuracy : 0.8989
## 95% CI : (0.8167, 0.9527)
## No Information Rate : 0.573
## P-Value [Acc > NIR] : 1.584e-11
##
## Kappa : 0.794
## Mcnemar's Test P-Value : 1
##
## Sensitivity : 0.8947
## Specificity : 0.9020
## Pos Pred Value : 0.8718
## Neg Pred Value : 0.9200
## Prevalence : 0.4270
## Detection Rate : 0.3820
## Detection Prevalence : 0.4382
## Balanced Accuracy : 0.8983
##
## 'Positive' Class : 0
## b.knn.confit(b.train20,b.test20)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 41 1
## 1 6 54
##
## Accuracy : 0.9314
## 95% CI : (0.8637, 0.972)
## No Information Rate : 0.5392
## P-Value [Acc > NIR] : <2e-16
##
## Kappa : 0.8608
## Mcnemar's Test P-Value : 0.1306
##
## Sensitivity : 0.8723
## Specificity : 0.9818
## Pos Pred Value : 0.9762
## Neg Pred Value : 0.9000
## Prevalence : 0.4608
## Detection Rate : 0.4020
## Detection Prevalence : 0.4118
## Balanced Accuracy : 0.9271
##
## 'Positive' Class : 0
## b.knn.confit(b.train21,b.test21)## Confusion Matrix and Statistics
##
## Reference
## Prediction 0 1
## 0 64 6
## 1 8 74
##
## Accuracy : 0.9079
## 95% CI : (0.8503, 0.9487)
## No Information Rate : 0.5263
## P-Value [Acc > NIR] : <2e-16
##
## Kappa : 0.815
## Mcnemar's Test P-Value : 0.7893
##
## Sensitivity : 0.8889
## Specificity : 0.9250
## Pos Pred Value : 0.9143
## Neg Pred Value : 0.9024
## Prevalence : 0.4737
## Detection Rate : 0.4211
## Detection Prevalence : 0.4605
## Balanced Accuracy : 0.9069
##
## 'Positive' Class : 0
##