SAS Enterprise Miner

RCLIB.DIABVA

Input Data Settings:

Source Data: RCLIB.DIABVA ( 51 rows, 7 columns)
Output: EMDATA.VIEW_PCW
Description: RCLIB.DIABVA
Role: VALIDATE
Metadata Sample: EMPROJ.SMP_VIOX ( 50 rows)

RCLIB.DIABTR

Input Data Settings:

Source Data: RCLIB.DIABTR ( 101 rows, 7 columns)
Output: EMDATA.VIEW_5RG
Description: RCLIB.DIABTR
Role: TRAIN
Metadata Sample: EMPROJ.SMP_VIZ8 ( 100 rows)

Neural Network

Optimization plot:

 
 Fit Statistic                     Training    Validation       Test 
  
 [ TARGET=DIAB ]                       .            .           . 
 Average Profit                       0.390        0.280       0.300 
 Misclassification Rate               0.240        0.200       0.240 
 Average Error                        0.504        0.485       0.517 
 Average Squared Error                0.166        0.157       0.167 
 Sum of Squared Errors               33.169       15.663      16.728 
 Root Average Squared Error           0.407        0.396       0.409 
 Root Final Prediction Error          0.441         .           . 
 Root Mean Squared Error              0.425        0.396       0.409 
 Error Function                     100.763       48.539      51.666 
 Mean Squared Error                   0.180        0.157       0.167 
 Maximum Absolute Error               0.807        0.813       0.923 
 Final Prediction Error               0.195         .           . 
 Divisor for ASE                    200.000      100.000     100.000 
 Model Degrees of Freedom             8.000         .           . 
 Degrees of Freedom for Error        92.000         .           . 
 Total Degrees of Freedom           100.000         .           . 
 Sum of Frequencies                 100.000       50.000      50.000 
 Sum Case Weights * Frequencies     200.000      100.000     100.000 
 Akaike's Information Criterion     116.763         .           . 
 Schwarz's Baysian Criterion        137.604         .           . 
  

• Objective function: Maximize profit

• Assessment Matrix: Profit vector

• Utilities

Confusion Matrix (Assessed Partition=VALIDATION)

End Report

Path Information

Name: Neural Network_T3VM7YMG
Target: DIAB
Description:
Mining Function: Transform
Subject: No subject
Rating: 0