Cluster Analysis - Profiling and Scoring

Cluster Analysis - Profiling and Scoring - Profiling...

Info iconThis preview shows pages 1–4. Sign up to view the full content.

View Full Document Right Arrow Icon
1 Profiling Clusters & Scoring New Observations Cluster Profiling Profiling can be defined as the generation of descriptions of the derived clusters from the input variables. • These descriptions are the class label for a cluster. • At least two (related) varieties of profiling exist: – Comparing the derived cluster means. – Comparing the derived means against a class.
Background image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
2 Comparing the derived cluster means 1. Generate a variable indicating to which cluster each observation belongs. 2. Call PROC MEANS to generate the centroids that define each cluster. 3. Perform ANOVA or MANOVA on the input variables to test for significant differences between the cluster centroids. MANOVA • If the multivariate tests are found to be significant, it is still not known which variables serve as the basis for the difference. • If the multivariate tests are found not to be significant, the groups may still significantly differ on the individual variables that define them.
Background image of page 2
3 Compare Known to Derived Clusters
Background image of page 3

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full DocumentRight Arrow Icon
Image of page 4
This is the end of the preview. Sign up to access the rest of the document.

This note was uploaded on 03/15/2010 for the course STATISTIC 472 taught by Professor Amjad during the Spring '08 term at Yarmouk University.

Page1 / 5

Cluster Analysis - Profiling and Scoring - Profiling...

This preview shows document pages 1 - 4. Sign up to view the full document.

View Full Document Right Arrow Icon
Ask a homework question - tutors are online