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 18 Clustering Data 477<br />

Self Organizing Maps<br />

So, in the limit as k becomes arbitrarily large, Φ k<br />

( x) tends toward the normal PDF. As k → 0 , Φ k<br />

( x)<br />

tends<br />

toward the exponential (Laplace) distribution.<br />

The regularization parameter k is set so that P(Normal(x) < k) = Huber Coverage, where Normal(x)<br />

indicates a multivariate normal variate. Huber Coverage is a user field, which defaults to 0.90.<br />

Self Organizing Maps<br />

The Self-Organizing Maps (SOMs) technique was developed by Teuvo Kohonen (1989) <strong>and</strong> further<br />

extended by a number of other neural network enthusiasts <strong>and</strong> statisticians. The original SOM was cast as a<br />

learning process, like the original neural net algorithms, but the version implemented here is done in a<br />

much more straightforward way as a simple variation on k-means clustering. In the SOM literature, this<br />

would be called a batch algorithm using a locally weighted linear smoother.<br />

The goal of a SOM is to not only form clusters, but form them in a particular layout on a cluster grid, such<br />

that points in clusters that are near each other in the SOM grid are also near each other in multivariate<br />

space. In classical k-means clustering, the structure of the clusters is arbitrary, but in SOMs the clusters have<br />

the grid structure. This grid structure helps to interpret the clusters in two dimensions: clusters that are close<br />

are more similar than distant clusters.<br />

To create a Self Organizing Map, select that option on the Method menu of the Iterative Clustering Control<br />

Panel (Figure 18.5). After selecting Self Organizing Map, the control panel looks like Figure 18.10.<br />

Figure 18.10 Self Organizing Map Control Panel<br />

Some of the options on the panel are described in “K-Means Control Panel” on page 470. The other options<br />

are described below:<br />

N Rows<br />

is the number of rows in the cluster grid.<br />

N Columns<br />

is the number of columns in the cluster grid.

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

Saved successfully!

Ooh no, something went wrong!