4 K-Nearest Neighbors

Brie

y describe the method in your own words. State any assumptions used and whether you think these assumptions are violated. Include relevant formula.

Use V -fold cross-validation to choose a value of k.

Plot the cross-validation error against k.

State the value of k chosen and the cross-validation error for this k.

5 QDA, LDA and FDA

Brie

y describe the methods in your own words. State any assumptions used and whether you think these assumptions are violated. Include relevant formula.

Find the V -fold cross-validation error for QDA, LDA and FDA.

Use the LDA and FDA coefficients to assess which variables increase the chances of diabetes.

6 Classification Trees

Briefly describe the method in your own words. State any assumptions used and whether you

think these assumptions are violated. Include relevant formula.

Find the V -fold cross-validation error for the classification tree t using the default settings in rpart.

Interpret the fitted tree using all of the data. Interpret the importance of each variable and the role each variable plays in predicting diabetes.

7 Logistic Regression

Briefly describe the method in your own words. State any assumptions used and whether you think these assumptions are violated. Include relevant formula.

Use appropriate methods discussed in class to select a model.

Find the V -fold cross-validation error for this model.

Use the selected logistic regression model to interpret the role each variable plays in predicting diabetes.

**Subject Mathematics Statistics-R Programming**