61 BIAS_H1 1.621410 3.94302E-7 BIAS -> H1
62 BIAS_H2 1.544387 -0.0000555 BIAS -> H2
63 BIAS_H3 -0.534732 0.0000123 BIAS -> H3
64 BIAS_H4 -0.785714 -0.0000101 BIAS -> H4
65 BIAS_H5 0.912585 -0.0000115 BIAS -> H5
66 BIAS_H6 -0.768781 -2.2982E-6 BIAS -> H6
67 BIAS_H7 1.015622 -9.1444E-6 BIAS -> H7
68 BIAS_H8 0.550634 -1.1615E-8 BIAS -> H8
69 BIAS_H9 -2.030613 -0.0000519 BIAS -> H9
70 BIAS_H10 0.306667 -9.3574E-6 BIAS -> H10
71 BIAS_H11 -2.591558 0.0000337 BIAS -> H11
72 BIAS_H12 1.854141 0.0000213 BIAS -> H12
73 BIAS_H13 -0.378726 -0.0000290 BIAS -> H13
74 BIAS_H14 -0.596093 -0.0000340 BIAS -> H14
75 BIAS_H15 1.200441 -0.0000417 BIAS -> H15
76 BIAS_H16 -1.358588 0.0000232 BIAS -> H16
77 BIAS_H17 -2.473932 0.0000370 BIAS -> H17
78 BIAS_H18 -1.391321 0.0000284 BIAS -> H18
79 BIAS_H19 -1.537930 -0.0000246 BIAS -> H19
80 BIAS_H20 0.717883 -0.0000180 BIAS -> H20
81 BIAS_H21 0.046762 -0.0000347 BIAS -> H21
82 BIAS_H22 -0.062103 -0.0000241 BIAS -> H22
83 BIAS_H23 0.090229 0.0000343 BIAS -> H23
84 BIAS_H24 0.502362 1.47214E-6 BIAS -> H24
85 BIAS_H25 -2.014471 -3.7539E-6 BIAS -> H25
86 BIAS_H26 -0.448834 7.76322E-6 BIAS -> H26
87 BIAS_H27 2.068868 0.0000197 BIAS -> H27
88 BIAS_H28 1.888770 0.0000405 BIAS -> H28
89 BIAS_H29 0.249710 -6.3913E-6 BIAS -> H29
90 BIAS_H30 3.409735 8.33278E-6 BIAS -> H30
91 H1_HIPL -0.110538 0.0000178 H1 -> HIPL
92 H2_HIPL -0.473553 -0.0000117 H2 -> HIPL
93 H3_HIPL -1.150299 0.0000209 H3 -> HIPL
94 H4_HIPL 0.477358 -4.1644E-6 H4 -> HIPL
95 H5_HIPL 0.464599 0.0000156 H5 -> HIPL
96 H6_HIPL 0.868824 0.0000395 H6 -> HIPL
97 H7_HIPL -0.305023 2.29377E-6 H7 -> HIPL
98 H8_HIPL -0.022398 0.0000134 H8 -> HIPL
99 H9_HIPL -0.971155 -0.0000579 H9 -> HIPL
100 H10_HIPL 0.974106 -0.0000192 H10 -> HIPL
Optimization Results
Parameter Estimates
------------------------------------------------------
Parameter Estimate Gradient Label
---------------------------------------------------
101 H11_HIPL 0.568802 -0.0000519 H11 -> HIPL
102 H12_HIPL 1.227553 0.0000166 H12 -> HIPL
103 H13_HIPL -0.466255 -0.0000231 H13 -> HIPL
104 H14_HIPL -0.894798 -0.0000322 H14 -> HIPL
105 H15_HIPL -1.479547 -2.2118E-6 H15 -> HIPL
106 H16_HIPL 0.471993 0.0000190 H16 -> HIPL
107 H17_HIPL -0.695108 -0.0000138 H17 -> HIPL
108 H18_HIPL 0.411108 -9.7179E-7 H18 -> HIPL
109 H19_HIPL -0.650073 2.01526E-7 H19 -> HIPL
110 H20_HIPL -1.728201 0.0000190 H20 -> HIPL
111 H21_HIPL -0.845374 0.0000182 H21 -> HIPL
112 H22_HIPL 0.907477 -0.0000158 H22 -> HIPL
113 H23_HIPL 0.890942 0.0000194 H23 -> HIPL
114 H24_HIPL 0.670337 -0.0000501 H24 -> HIPL
115 H25_HIPL 0.524875 -0.0000589 H25 -> HIPL
116 H26_HIPL 0.644267 -0.0000242 H26 -> HIPL
117 H27_HIPL -0.968536 -4.3518E-6 H27 -> HIPL
118 H28_HIPL -0.283694 1.1348E-6 H28 -> HIPL
119 H29_HIPL 0.366848 -2.7894E-6 H29 -> HIPL
120 H30_HIPL 0.731690 0.0000279 H30 -> HIPL
121 BIAS_HIP -0.008245 0.0000135 BIAS -> HIPL
Value of Objective Function = 0.0007488987
PROC PRINT Report of the Average Squared Error for the Scored Test Data Set
Hill & Plateau Data
MLP with 30 Hidden Units
Fit Statistics for the Test Data
Test:
Average
Squared
Error.
.00071717
GCONTOUR Plot of the Predicted Values
G3D Plot of the Predicted Values
Copyright 2000 by SAS Institute Inc., Cary, NC, USA. All rights reserved.
The %LET statement sets the macro variable HIDDEN to 3.
title 'Hill & Plateau Data';
%let hidden=3;
The MAXITER = option specifies the maximum number of iterations.
hidden &hidden / id=h;
prelim 10;
train maxiter=1000 outest=mlpest;
score data=sampsio.dmtsurf out=mlpout outfit=mlpfit;
title2 "MLP with &hidden Hidden Units";
run;
PROC PRINT creates a report of selected fit statistics.
proc print data=mlpfit noobs label;
var _tase_ _tasel_ _taseu_;
where _name_ ='HIPL';
title3 'Fit Statistics for the Test Data';
run;
PROC GCONTOUR creates a plot of the predicted values.
proc gcontour data=mlpout;
plot x2*x1=p_hipl / pattern ctext=black coutline=gray;
pattern v=msolid;
legend frame;
title3 'Predicted Values';
footnote;
run;