Make a presentation of you doing the follows. Using Zoom to record a video of the presentation and the link of the video or the video to Canvas.

Video Instruction

  1. Download and Import to SAS Miner the Breast Cancer Wisconsin (Diagnostic) Data Set at this link.

  2. Reject all the text variables. Set diagnosis as the binary target.

  3. Split the data 70:30:0 for Training: Validation: Test. Create the following decision trees.

    • a decision tree used Gini Index for splitting (Nominal Target Criterion)

    • a decision tree used Entropy for splitting (Nominal Target Criterion)

    • a decision tree used Chi-squared test for splitting (Nominal Target Criterion)

    • a decision tree with 4 leaves.

  4. Compare the trees in terms of misclassification rate and ROC-Index.

  5. Show the confusion matrix (classification table) of the Entropy tree.

  6. What is the best tree in term of misclassification. In that tree, what is the most important variables?

  7. How many leaves that tree in 6. have? Explain why the tree ended up with that number of leaves. (i.e., explain the Subtree Assessment Plot).