Book a Demo!
CoCalc Logo Icon
StoreFeaturesDocsShareSupportNewsAboutPoliciesSign UpSign In
debakarr
GitHub Repository: debakarr/machinelearning
 Path: blob/master/Part 10 - Model Selection And Boosting/k-Fold Cross Validation/k_fold_cross_validation.R
1341 views
1
# k-Fold Cross Validation

2


3
# Importing the dataset

4
dataset = read.csv('Social_Network_Ads.csv')

5
dataset = dataset[3:5]

6


7
# Encoding the target feature as factor

8
dataset$Purchased = factor(dataset$Purchased, levels = c(0, 1))

9


10
# Splitting the dataset into the Training set and Test set

11
# install.packages('caTools')

12
library(caTools)

13
set.seed(123)

14
split = sample.split(dataset$Purchased, SplitRatio = 0.75)

15
training_set = subset(dataset, split == TRUE)

16
test_set = subset(dataset, split == FALSE)

17


18
# Feature Scaling

19
training_set[-3] = scale(training_set[-3])

20
test_set[-3] = scale(test_set[-3])

21


22
# Fitting Kernel SVM to the Training set

23
# install.packages('e1071')

24
library(e1071)

25
classifier = svm(formula = Purchased ~ .,

26
                 data = training_set,

27
                 type = 'C-classification',

28
                 kernel = 'radial')

29


30
# Predicting the Test set results

31
y_pred = predict(classifier, newdata = test_set[-3])

32


33
# Making the Confusion Matrix

34
cm = table(test_set[, 3], y_pred)

35


36
# Applying k-Fold Cross Validation

37
# install.packages('caret')

38
library(caret)

39
folds = createFolds(training_set$Purchased, k = 10)

40
cv = lapply(folds, function(x) {

41
  training_fold = training_set[-x, ]

42
  test_fold = training_set[x, ]

43
  classifier = svm(formula = Purchased ~ .,

44
                   data = training_fold,

45
                   type = 'C-classification',

46
                   kernel = 'radial')

47
  y_pred = predict(classifier, newdata = test_fold[-3])

48
  cm = table(test_fold[, 3], y_pred)

49
  accuracy = (cm[1,1] + cm[2,2]) / (cm[1,1] + cm[2,2] + cm[1,2] + cm[2,1])

50
  return(accuracy)

51
})

52
accuracy = mean(as.numeric(cv))

53


54
# Visualising the Training set results

55
library(ElemStatLearn)

56
set = training_set

57
X1 = seq(min(set[, 1]) - 1, max(set[, 1]) + 1, by = 0.01)

58
X2 = seq(min(set[, 2]) - 1, max(set[, 2]) + 1, by = 0.01)

59
grid_set = expand.grid(X1, X2)

60
colnames(grid_set) = c('Age', 'EstimatedSalary')

61
y_grid = predict(classifier, newdata = grid_set)

62
plot(set[, -3],

63
     main = 'Kernel SVM (Training set)',

64
     xlab = 'Age', ylab = 'Estimated Salary',

65
     xlim = range(X1), ylim = range(X2))

66
contour(X1, X2, matrix(as.numeric(y_grid), length(X1), length(X2)), add = TRUE)

67
points(grid_set, pch = '.', col = ifelse(y_grid == 1, 'springgreen3', 'tomato'))

68
points(set, pch = 21, bg = ifelse(set[, 3] == 1, 'green4', 'red3'))

69


70
# Visualising the Test set results

71
library(ElemStatLearn)

72
set = test_set

73
X1 = seq(min(set[, 1]) - 1, max(set[, 1]) + 1, by = 0.01)

74
X2 = seq(min(set[, 2]) - 1, max(set[, 2]) + 1, by = 0.01)

75
grid_set = expand.grid(X1, X2)

76
colnames(grid_set) = c('Age', 'EstimatedSalary')

77
y_grid = predict(classifier, newdata = grid_set)

78
plot(set[, -3], main = 'Kernel SVM (Test set)',

79
     xlab = 'Age', ylab = 'Estimated Salary',

80
     xlim = range(X1), ylim = range(X2))

81
contour(X1, X2, matrix(as.numeric(y_grid), length(X1), length(X2)), add = TRUE)

82
points(grid_set, pch = '.', col = ifelse(y_grid == 1, 'springgreen3', 'tomato'))

83
points(set, pch = 21, bg = ifelse(set[, 3] == 1, 'green4', 'red3'))

84