Practical Machine Learning Week 3

library(tidyverse)
library(caret)
library(rattle)
data("iris")

table(iris$Species)

    setosa versicolor  virginica 
        50         50         50 

inTrain <- createDataPartition(iris$Species, p = 0.7, list = FALSE)
training <- iris[inTrain, ]
testing <- iris[-inTrain, ]

qplot(Petal.Width, Sepal.Width, col = Species, data = training)

modFit <- train(Species ~., method = "rpart", data = training)
modFit$finalModel

n= 105 

node), split, n, loss, yval, (yprob)
      * denotes terminal node

1) root 105 70 setosa (0.33333333 0.33333333 0.33333333)  
  2) Petal.Length< 2.45 35  0 setosa (1.00000000 0.00000000 0.00000000) *
  3) Petal.Length>=2.45 70 35 versicolor (0.00000000 0.50000000 0.50000000)  
    6) Petal.Width< 1.75 36  2 versicolor (0.00000000 0.94444444 0.05555556) *
    7) Petal.Width>=1.75 34  1 virginica (0.00000000 0.02941176 0.97058824) *

plot(modFit$finalModel, uniform = TRUE, main = "Classification Tree")
text(modFit$finalModel, use.n = TRUE, all = TRUE, cex = 0.8 )

fancyRpartPlot(modFit$finalModel, main = "Classification Tree")

predict(modFit, newdata = testing)

 [1] setosa     setosa     setosa     setosa     setosa     setosa     setosa     setosa     setosa     setosa     setosa    
[12] setosa     setosa     setosa     setosa     versicolor versicolor versicolor versicolor versicolor versicolor versicolor
[23] versicolor versicolor versicolor versicolor versicolor versicolor versicolor versicolor versicolor virginica  virginica 
[34] virginica  virginica  virginica  virginica  virginica  virginica  versicolor virginica  virginica  versicolor virginica 
[45] virginica 
Levels: setosa versicolor virginica

Bootstrap aggregating(bagging)

data(ozone, package = "ElemStatLearn")
ozone <- ozone[order(ozone$ozone), ]
head(ozone)

ll <- matrix(NA, nrow = 10, ncol = 155)
for (i in 1:10) {
  ss <- sample(1:dim(ozone)[1], replace = TRUE)
  ozone0 <- ozone[ss, ]
  ozone0 <- ozone0[order(ozone0$ozone), ]
  loess0 <- loess(temperature ~ ozone, data = ozone0, span = 0.2)
  ll[i, ] <- predict(loess0, newdata = data.frame(ozone = 1:155))
}

plot(ozone$ozone, ozone$temperature, pch = 19, cex = 0.5)
for (i in 1:10) {
  lines(1:155, ll[i, ], col ="grey", lwd = 2)
}

lines(1:155, apply(ll, 2, mean), col = "red", lwd = 2)

library(party)
predictors <- data.frame(ozone=ozone$ozone)
temperature = ozone$temperature
treebag <- bag(predictors, temperature, B= 10,
               bagControl = bagControl(fit = ctreeBag$fit,
                                       predict = ctreeBag$pred,
                                       aggregate = ctreeBag$aggregate))

plot(ozone$ozone, temperature, col = "lightgrey", pch = 19)
points(ozone$ozone, predict(treebag$fits[[1]]$fit, predictors), pch = 19, col = "red")

points(ozone$ozone, predict(treebag, predictors), pch= 19, col = "blue")

NA
NA

Random Forests

modFit

Random Forest 

105 samples
  4 predictor
  3 classes: 'setosa', 'versicolor', 'virginica' 

No pre-processing
Resampling: Bootstrapped (25 reps) 
Summary of sample sizes: 105, 105, 105, 105, 105, 105, ... 
Resampling results across tuning parameters:

  mtry  Accuracy   Kappa    
  2     0.9440576  0.9152679
  3     0.9430061  0.9137014
  4     0.9407163  0.9102969

Accuracy was used to select the optimal model using the largest value.
The final value used for the model was mtry = 2.

getTree(modFit$finalModel, k=2)

   left daughter right daughter split var split point status prediction
1              2              3         3        2.60      1          0
2              0              0         0        0.00     -1          1
3              4              5         1        6.15      1          0
4              6              7         3        4.75      1          0
5              8              9         3        5.05      1          0
6              0              0         0        0.00     -1          2
7              0              0         0        0.00     -1          3
8             10             11         1        6.40      1          0
9              0              0         0        0.00     -1          3
10            12             13         4        1.65      1          0
11             0              0         0        0.00     -1          2
12             0              0         0        0.00     -1          2
13             0              0         0        0.00     -1          3

irisP <- classCenter(training[ ,c(3,4)], training$Species, modFit$finalModel$proximity)
irisP <- as.data.frame(irisP)
irisP$Species <- rownames(irisP)
p <- qplot(Petal.Width, Petal.Length, col = Species, data = training)
p + geom_point(aes(x = Petal.Width, y= Petal.Length, col = Species), size = 5, shape = 14, data = irisP)

table(pred, testing$Species)

            
pred         setosa versicolor virginica
  setosa         15          0         0
  versicolor      0         15         3
  virginica       0          0        12

qplot(Petal.Width, Petal.Length, col = predRight, data = testing, main = "newdata Predictions")

Boosting

library(ISLR)
data("Wage")
inTrain2 <- createDataPartition(y = Wage$wage,
                               p = 0.7, list = FALSE)
training2 <- Wage[inTrain2, ]
testing2 <-  Wage[-inTrain2, ]

modFit2 <- train(wage ~., method = "gbm", data = training2, verbose = FALSE)
modFit2

## plot the result
qplot(predict(modFit2, testing2), wage, data = testing2)

Model Based Prediction

modlda <- train(Species ~., method = "lda", data = training)
modnb <- train(Species ~., method = "nb", data = training)
plda <- predict(modlda, testing)
pnb <- predict(modnb, testing)
table(plda, pnb)

            pnb
plda         setosa versicolor virginica
  setosa         15          0         0
  versicolor      0         14         1
  virginica       0          2        13

equalPredictions = (plda == pnb)
qplot(Petal.Width, Sepal.Width, col = equalPredictions, data = testing, main = "Comparison of Result")