Question

R project for building a k-NN model. In this question you are going to work with perhaps the best known dataset in the machine learning community. It is about the classification of the iris flower in different species. The original dataset contains 4 attributes and 3 classes, but, to simplify the problem, we are going to work with a subset of the data. The dataset we are going to work with is in iris.csv, and contains only two (continuous) features: the petal length in cm, and the petal width in cm; and two classes: iris versicolor, and iris verginica.

Use the repeated holdout technique to find the near optimal k for the k-NN method for this data. To this end use values of k = 1 to k = 49 with jumps of 4, that is, test k = 1, 5, 9, . . . , 49. For each k run 20 experiments where you would choose 80 random items from the data as your training set, and the remaining items as your test set. For each run, build a k-NN model, and test it on the test set, and find number of misclassified versicolor items, misclassified verginica items and the total number of misclassified items. Take the average over 20 experiments, and divide, respectively by the number of versicolor items, the number of verginica items, and the total number of items in the test set. Collect these three items, along with the values of k in vectors. When done find the best k, that is the one resulting in lowest error rate. Also on the same graphics panel, graph versicolor error rate (proportion of versicolor points incorrectly classified as verginica among all orange points), verginica error rate (proportion of verginica points incorrectly misclassified as versicolor ) and total error rate against values of k.

Answer 1

#Loading package

install.packages('ggplot2')

#Loading Data

#In order to load data Internet is required.

data = read.csv(url("http://archive.ics.uci.edu/ml/machine-learning-databases/iris/iris.data"), header = FALSE)

head(iris)

#Dividing into 80-20 percent

ind <- sample(2, nrow(iris), replace=TRUE, prob=c(0.80, 0.20))
table(ind)

#Assigning Variables

iris_train <- iris[ind==1, 1:4]
iris_test <- iris[ind==2, 1:4]
iris_train_labels <- iris[ind==1, 5]
iris_test_labels <- iris[ind==2, 5]

library(class)

#Importing Knn function
iris_test_pred<-knn(train=iris_train, test=iris_test,cl=iris_train_labels,k=3)

library(gmodels)
CrossTable(x=iris_test_labels, y=iris_test_pred, prop.chisq = FALSE)