14.03.2014 Views

Modeling and Multivariate Methods - SAS

Modeling and Multivariate Methods - SAS

Modeling and Multivariate Methods - SAS

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Chapter 13 Recursively Partitioning Data 317<br />

Introduction to Partitioning<br />

Introduction to Partitioning<br />

Variations of partitioning go by many names <strong>and</strong> br<strong>and</strong> names: decision trees, CART TM , CHAID TM , C4.5,<br />

C5, <strong>and</strong> others. The technique is often taught as a data mining technique because<br />

• it is good for exploring relationships without having a good prior model,<br />

• it h<strong>and</strong>les large problems easily, <strong>and</strong><br />

• the results are very interpretable.<br />

A classic application is where you want to turn a data table of symptoms <strong>and</strong> diagnoses of a certain illness<br />

into a hierarchy of questions to ask new patients in order to make a quick initial diagnosis.<br />

The factor columns (X’s) can be either continuous or categorical (nominal or ordinal). If an X is continuous,<br />

then the splits (partitions) are created by a cutting value. The sample is divided into values below <strong>and</strong> above<br />

this cutting value. If the X is categorical, then the sample is divided into two groups of levels.<br />

The response column (Y) can also be either continuous or categorical (nominal or ordinal). If Y is<br />

continuous, then the platform fits means. If Y is categorical, then the fitted value is a probability. In either<br />

case, the split is chosen to maximize the difference in the responses between the two branches of the split.<br />

For more information on split criteria, see “Statistical Details” on page 348.<br />

Launching the Partition Platform<br />

To launch the Partition platform, select Analyze > <strong>Modeling</strong> > Partition. The Partition launch window is<br />

shown in Figure 13.2, using the Boston Housing.jmp data table.<br />

Figure 13.2 Partition Launch Window

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!