DSC 410/510 - Multivariate Statistical Methods

Assignment 9 (deadline: 4pm on Monday, November 1)

Practical matters

Class demographics example:

• Work through "discrim1.sas" and "discrim2.sas" (available from the handouts page of the class web-site) which provides SAS code and instructions for running discriminant analyses on the class demographics data (available in the "discrim.xls" spreadsheet on the data page).
• Send an e-mail to ipardoe at lcbmail.uoregon.edu with the following information:
  1. For the two group example (discrim1.sas), five people are misclassified by the procedure. Tell me:
     1. the ID numbers of the three people that were in cluster 1, but are classified by the procedure into cluster 2
     2. the ID numbers of the two people that were in cluster 2, but are classified by the procedure into cluster 1
  2. For the three group example (discrim2.sas), one person is misclassified by the procedure using Age and Salary. Does this misclassification change when re-running the analysis using just the first canonical discriminant function: yes or no?

Reading

Chapter 5 in the text-book, pages 239-325.