Use Liver Disorders Data Set. Make sure you identify the two groups in Variable 7 correctly as group 1 ‘severe’ group 2 ‘not severe’
a. Use LDA to classify observations (people) into the two groups. Assume prior probabilities based on the available sample (dataset)
b.Minimize the total error of misclassification by changing the prior probabilities for the two populations. Show the results graphically (error rates vs. prior on the severe group), and also show the optimal values, including the misclassification table for that optimal case.
c. Repeat Steps 1 and 2 with 3-fold CV; use 100 rounds. Calculate the standard error of the minimized total error of misclassification.
d.Repeat Steps 1 and 2 with 10-fold CV; use 100 rounds. Calculate the standard error of the minimized total error of misclassification.
0 comments