## Maximum Redundancy Analysis

In contrast to canonical redundancy analysis, which
examines how well the original variables can be
predicted from the canonical variables,
maximum redundancy analysis finds the variables
that can best predict the original variables.
Given two sets of variables, maximum redundancy analysis
finds a linear combination from one set of variables that
best predicts the variables in the opposite set.
SAS/INSIGHT software normalizes the coefficients of the
linear combinations so that each maximum redundancy variable
has a variance of 1.

Maximum redundancy analysis continues by finding a second
maximum redundancy variable from one set of variables,
uncorrelated with the first one,
that produces the second highest prediction power
for the variables in the opposite set. The process of constructing
maximum redundancy variables continues until the number
of maximum redundancy variables equals the number of
variables in the smaller group.

Either raw variances (**Raw Variance**) or standardized variances
(**Std Variance**) can be used in the analysis.
You specify the selection in the method
options dialog as shown in Figure 40.3.
By default, standardized variances are used.

Copyright © 1999 by SAS Institute Inc., Cary, NC, USA. All rights reserved.