DMREG Response Profile
The Response Profile table lists the target categories, their ordered values, and their total frequencies for the given data.
DMREG Optimization Table
The Optimization table provides a summary of the Newton-Raphson Ridge optimization results.
DMREG Model Fitting Information and Testing Global Null Hypothesis Beta=0
The Model Fitting Information and Testing Global Null Hypothesis Beta=0 table contains the negative of twice the log likelihood (-2 LOG
L) for the fitted model. Results of the likelihood ratio test and the efficient score test for testing the joint significance of the explanatory
inputs are also printed in the table.
DMREG Analysis of Maximum Likelihood Estimates
The Analysis of Maximum Likelihood Estimates table lists the parameter estimates, their standard errors, and the results of the Wald test for
the individual parameters. A standardized estimate for each slope parameter and the odds ratio for each estimate is also printed. An odds ratio
is obtained by exponentiating the corresponding parameter estimate.
DMREG Odds Ratio Estimates
The Odd Ratio Estimates table lists the odd ratios for the explanatory inputs. The odd ratio estimates provide the change in odds for a unit
increase in each input.
PROC PRINT Report of Selected Fit Statistics for the Training Data Set
The misclassification rate for the training data set is only 37.22%.
PROC FREQ Misclassification Table for the Training Data
All observations in the training data are classified into the C=3 level. The linear model is not adequate.
PROC GPLOT Plot of the Classification Results
The target classes are not linearly separable.
Quadratic-Logistic Program
proc dmreg data=sampsio.dmdring dmdbcat=sampsio.dmdring;
class c;
model c=x|x|y|y @2;
score out=qout outfit=qfit;
score data=sampsio.dmsring nodmdb out=qgridout;
title1 'Quadratic-Logistic Regression with Ordinal Target';
run;
proc print data=qfit noobs label;
var _aic_ _max_ _rfpe_ _misc_;
title2 'Fit Statistics for the Training Data Set';
run;
proc freq data=qout;
tables f_c*i_c;
title2 'Misclassification Table: Training Data';
run;
proc gplot data=qout;
plot y*x=i_c / haxis=axis1 vaxis=axis2;
symbol c=black i=none v=dot;
symbol2 c=red i=none v=square;
symbol3 c=green i=none v=triangle;
axis1 c=black width=2.5 order=(0 to 30 by 5);
axis2 c=black width=2.5 minor=none order=(0 to 20 by 2);
title2 'Classification Results';
run;
proc gcontour data=qgridout;
plot y*x=p_c1 / pattern ctext=black coutline=gray;
plot y*x=p_c2 / pattern ctext=black coutline=gray;;
plot y*x=p_c3 / pattern ctext=black coutline=gray;;
title2 'Posterior Probabilities';
pattern v=msolid;
legend frame;
run;
Quadratic-Logistic Output
DMREG Output
PROC PRINT Report of Selected Fit Statistics for the Training Data
Note that the training misclassification rate is 0. All cases are correctly classified by the quadratic-logistic model.
PROC FREQ Misclassification Table for the Training Data
PROC GPLOT Plot of the Classification Results
PROC GCONTOUR Plots of the Posterior Probabilities
Copyright 2000 by SAS Institute Inc., Cary, NC, USA. All rights reserved.
PROC GPLOT creates a scatter plot of the Rings training data.
proc gplot data=sampsio.dmdring;
plot y*x=c /haxis=axis1 vaxis=axis2;
symbol c=black i=none v=dot;
symbol2 c=red i=none v=square;
symbol3 c=green i=none v=triangle;
axis1 c=black width=2.5 order=(0 to 30 by 5);
axis2 c=black width=2.5 minor=none order=(0 to 20 by 2);
title 'Plot of the Rings Training Data';
run;
The PROC DMREG statement invokes the procedure. The DATA= option identifies
the DMDB encoded training data set that is used to fit the model. The DMDBCAT=option
identifies the DMDB training data catalog. You can create DMDB encoded data
sets and catalogs with the DMDB procedure.
proc dmreg data=sampsio.dmdring dmdbcat=sampsio.dmdring;
The CLASS statement identifies the target C as a categorical variable.
class c;
The MODEL statement specifies the linear-logistic model.
model c = x y;
Dostları ilə paylaş: |