Cross Validation Activity Sheet The image below shows a

Cross Validation Activity Sheet

rn

The image below shows a scatterplot for a dataset with 6 people (A,B,C,D,E, and F). Each person has a score for two predictor variables and an outcome variable (Win vs. Lose).

Imagine that you performed a 6-fold cross-validation on the data. Fill in the 6 tables that would be created from the 6-fold cross-validation procedure. The first table (Fold 1) is already completed for you. Work in reverse alphabetical order, where the First fold’s test set has person F, the 2nd Fold’s test set has person E, the 3rd Fold’s test set has person D.

Fold 1:

Train Dataset

Test Dataset

Fold 2:

Train Dataset

Test Dataset

Fold 3:

Train Dataset

Test Dataset

Fold 4:

Train Dataset

Test Dataset

Fold 5:

Train Dataset

Test Dataset

Fold 6:

Train Dataset

Test Dataset

——————————————————————————————————–

If you used a 3-Nearest Neighbor Model to predict the test outcome from the training dataset, what is your error (accuracy) on the test set for….

*Hint: You can look at the image below to figure out the 3-nearest neighbors for the case you are predicting in each test set

Test Accuracy for Fold 1_________________

Test Accuracy for Fold 2________________

Test Accuracy for Fold 3________________

Test Accuracy for Fold 4________________

Test Accuracy for Fold 5________________

Test Accuracy for Fold 6________________

Based on the errors above, what is the average cross-validated accuracy when using a 3-Nearest neighbor Model__________

Create Tables for a 2-Fold Cross validation using the same image as before. You can choose which participants are assigned to go into the first fold’s training and test set randomly.
rn