1. Use the Pima.tr data set in MASS package. This dataset includes diabetes information about 532 women who were at least 21 years old of Pima Indian heritage and living near Phoenix, Arizona. > library (MASS) > data (Pima.tr) npreg glu bp skin bmi number of pregnancy plasma glucose concentration in an oral glucose tolerance test diastolic blood pressure (mm Hg) triceps skin fold thickness (mm) body mass index (weight in kg/(height in m)?) diabetes pedigree function age in year Yes or No, for diabetic according to WHO criteria ped age type (a) Fit a logistic regression model and discriminant analysis model. (b) Using a classification threshold of 50%, produce confusion matrices under logistic regression and LDA. What are the sensitivities and specificities for each model, and which model do you prefer? The information about the variables is as follows: (c) Produce ROC curves for A and B, and report each model's AUC. Which model seems to perform better on the training data? (d) Choose the "best" classification threshold (perhaps other than 50%) based on some criterion involving the TPR (true positive rate) and FPR (false positive rate). Would you use the same thresholds for both models? (e) Repeat (a) for your favorite classification threshold from part (c) 1. Use the Pima.tr data set in MASS package. This dataset includes diabetes information about 532 women who were at least 21 years old of Pima Indian heritage and living near Phoenix, Arizona. > library (MASS) > data (Pima.tr) npreg glu bp skin bmi number of pregnancy plasma glucose concentration in an oral glucose tolerance test diastolic blood pressure (mm Hg) triceps skin fold thickness (mm) body mass index (weight in kg/(height in m)?) diabetes pedigree function age in year Yes or No, for diabetic according to WHO criteria ped age type (a) Fit a logistic regression model and discriminant analysis model. (b) Using a classification threshold of 50%, produce confusion matrices under logistic regression and LDA. What are the sensitivities and specificities for each model, and which model do you prefer? The information about the variables is as follows: (c) Produce ROC curves for A and B, and report each model's AUC. Which model seems to perform better on the training data? (d) Choose the "best" classification threshold (perhaps other than 50%) based on some criterion involving the TPR (true positive rate) and FPR (false positive rate). Would you use the same thresholds for both models? (e) Repeat (a) for your favorite classification threshold from part (c)