PROC DMNEURL: Approximation to PROC NEURAL
One numeric (interval scaled) variable may be speciﬁed as a WEIGHT variable. It is
recommended to specify the WEIGHT variable already in the PROC DMDB invoca-
tion. Then the information is saved in the catalog and that variable is used automati-
cally as a FREQ variable in PROC DMNEURL.
Scoring the Model Using the OUTEST= Data set
The score value
is computed for each observation
value of the target (response) variable
of the input data set. All information needed
OUTEST= data set. First an observation from the input data set is mapped into a
new values in which
1. CLASS predictor variables with
categories are replaced by
dummy (binary) variables, depending on the fact whether the variable has miss-
2. Missing values in interval predictor variables are replaced by the mean value of
this variable in the DMDB data set. This mean value is taken from the catalog
of the DMDB data set.
3. The values of a WEIGHT or FREQ variable are multiplied into the observation.
4. For an interval target variable
its value is transformed into the interval [0,1]
5. All predictor variables are transformed into values with zero mean and unit
standard deviation by
are listed in the OUTEST= data set.
has more entries
The scoring is additive across the stages. The following information is available for
scoring each stage
each of dimension
the best activation function
optimal parameter estimates
similar to principal component analysis. With those values
the model can be ex-
is the speciﬁed link function.
In other words, this means, that given the
is computed from
are two of the
is deﬁned as
, and for the last component
The link function
is applied on
and yields to
are added to the predicted value (posterior)
PROC DMREG Statement
Example 1: Linear and Quadratic Logistic Regression with an Ordinal Target (Rings Data)
Example 2: Performing a Stepwise OLS Regression (DMREG Baseball Data)
Example 3: Comparison of the DMREG and LOGISTIC Procedures when Using a Categorical Input
Copyright 2000 by SAS Institute Inc., Cary, NC, USA. All rights reserved.