Problem 4 1) Consider the vowel training data, which has 11 classes with XR1n. Follow the procedure on page 114 to find the first and second discriminant variables (also called canonical coordinates). then draw the scatter plot and mark the centroids in the two dimensional subspace spanned by the first two canonical variates (see Figure 4.11). 2) Draw the decision boundary as in Figure 4.11 , and report the classilication accuracy. Problem 5 Lse linear regression and logistic regression to classify the vowel data in Problem 4, and report classification accuracy for each of them. Problem 4 1) Consider the vowel training data, which has 11 classes with XR1n. Follow the procedure on page 114 to find the first and second discriminant variables (also called canonical coordinates). then draw the scatter plot and mark the centroids in the two dimensional subspace spanned by the first two canonical variates (see Figure 4.11). 2) Draw the decision boundary as in Figure 4.11 , and report the classilication accuracy. Problem 5 Lse linear regression and logistic regression to classify the vowel data in Problem 4, and report classification accuracy for each of them