SAS Enterprise Miner

User : esposito
Date : 18JAN2005:16:21:37
Notes:


"EM Workspace" :

EM Workspace


PERSO.ZIPMONTEST

Input Data Settings:


  • All variables

  • Interval Variables

  • Notes: not available


    PERSO.ZIPTRAIN

    Input Data Settings:


  • All variables

  • Interval Variables

  • Notes: not available


    Memory-Based Reasoning (Exp.)

  • Settings

  • Variables

  • Output

  • Log

  • Training Code

  • Score Code

    Model assessment settings
    Train data set is selected for assessment.
    Validation data set is not selected for assessment.
    Test data set is not selected for assessment.
    Scored data set: 5000 observations are saved for interactive model assessment.


    Data Partition

  • Partition Settings

  • Output

  • Log

  • Training Code

  • Notes: not available


    Tree

    Model assessment plot:

    SAS Graphics

     
     Fit Statistic                     Training    Validation       Test 
      
     Average Squared Error                 1.94         3.00        3.72 
     Sum of Squared Errors              2717.74      1800.02     7448.34 
     Root Average Squared Error            1.39         1.73        1.93 
     Maximum Absolute Error                6.09         7.93        8.93 
     Divisor for ASE                    1400.00       600.00     2000.00 
     Total Degrees of Freedom           1400.00          .           . 
     Number of Estimated Weights          29.00          .           . 
     Sum of Frequencies                 1400.00       600.00     2000.00 
     Sum Case Weights * Frequencies     1400.00       600.00     2000.00 
      
     
      LEAF                    ROOT      V ROOT                  V 
        ID      N    V N      ASE        ASE      AVERAGE    AVERAGE 
      
        58    121     58    1.34139    2.63980    1.35537    2.36207 
        59     24      8    2.25924    1.54110    4.75000    4.87500 
        33     98     48    0.41033    1.25323    0.07143    0.27083 
        17     15      2    1.23648    2.20706    5.93333    5.00000 
        18     25     12    1.38564    1.74452    5.40000    4.41667 
        19      3      1    0.94281    0.33333    1.66667    2.00000 
        20    146     76    1.57320    1.72187    3.41096    3.18421 
        38     32     13    0.00000    0.00000    1.00000    1.00000 
        39      4      0    0.00000     .         2.00000     . 
        22    130     49    2.13918    2.06314    6.09231    6.20408 
        23     66     35    1.80735    2.16698    4.22727    4.65714 
        74     34     11    1.39202    1.57800    5.05882    4.72727 
        75     56     19    1.61457    1.53203    7.51786    7.73684 
        76     97     39    1.72001    1.67046    5.10309    4.94872 
        77     16      3    2.07572    1.19188    1.93750    2.66667 
        46     62     14    0.39639    1.08276    2.06452    2.35714 
        47     27     13    1.75291    2.14269    5.03704    4.30769 
        48     36     19    0.00000    0.97333    1.00000    1.31579 
        49      3      2    0.00000    6.00000    2.00000    8.00000 
        27     12      8    0.95379    2.83701    3.08333    4.25000 
        82     42     16    0.00000    0.00000    4.00000    4.00000 
        83      1      1    0.00000    1.00000    8.00000    7.00000 
        51      8      4    0.85696    0.97628    6.62500    5.75000 
        52     40     14    0.75788    0.87459    5.77500    6.00000 
        53      6      1    0.74536    1.33333    7.66667    9.00000 
        54     92     40    1.22103    1.15385    6.85870    6.77500 
        88    156     67    0.81254    1.20054    8.49359    8.28358 
        89      3      1    0.47140    0.33333    5.33333    5.00000 
        31     45     26    1.95050    1.68751    5.20000    5.57692 
      
  • English rules

  • Sequence

  • Matrix

    Target information
    Name: VAR257
    Label:
    Measurement: interval

    Tree settings

    No profile information defined
    Splitting criterion: F Test
    Significance Level: 0.2
    Minimum number of observations in a leaf: 1
    Observations required for a split search: 20
    Maximum number of branches from a node: 2
    Maximum depth of tree: 6
    Splitting rules saved in each node: 5
    Surrogate rules saved in each node: 0
    Treat missing as an acceptable value
    Model assessment measure: Average Squared Error
    Subtree: Best assessment value
    Observations sufficient for split search: 1401
    Maximum tries in an exhaustive split search: 5000
    P-value adjustment: KASS DEPTH
    Apply KASS BEFORE choosing number of branches

  • Log

  • Score Code
    Model assessment settings
    Train data set is not selected for assessment.
    Validation data set is selected for assessment.
    Test data set is not selected for assessment.
    Scored data set: 5000 observations are saved for interactive model assessment.

  • Notes: not available


    Tree

    Model assessment plot:

    SAS Graphics

     
     Fit Statistic                     Training    Validation       Test 
      
     Average Squared Error                 1.47         .           3.37 
     Sum of Squared Errors              2930.04         .        6741.51 
     Root Average Squared Error            1.21         .           1.84 
     Maximum Absolute Error                6.37         .           8.81 
     Divisor for ASE                    2000.00         .        2000.00 
     Total Degrees of Freedom           2000.00         .            . 
     Number of Estimated Weights          54.00         .            . 
     Sum of Frequencies                 2000.00         .        2000.00 
     Sum Case Weights * Frequencies     2000.00         .        2000.00 
      
     
      LEAF             ROOT 
        ID      N      ASE      AVERAGE 
      
        60     50    2.06068    3.44000 
        61     28    0.89500    0.35714 
        62      9    2.23331    3.11111 
        63     20    0.77460    8.00000 
        64    215    1.73102    1.62791 
        65    133    0.67334    0.18797 
        66     21    0.92091    5.09524 
        67      2    0.00000    0.00000 
        68     20    0.77460    4.00000 
        69      5    0.63246    2.00000 
        37      2    1.00000    7.00000 
        70     18    0.31427    5.88889 
        71      2    0.50000    8.50000 
        72     21    0.39268    4.80952 
        73      1    0.00000    2.00000 
        74     49    1.44452    4.48980 
        75      4    0.43301    1.25000 
        76    115    0.81711    3.04348 
        77     23    0.49573    1.56522 
        21      3    0.47140    8.33333 
        78     62    1.82885    5.24194 
        79     17    1.33362    7.52941 
        80     46    1.16035    7.84783 
        81      4    1.29904    3.75000 
        82     24    0.81543    2.79167 
        83      1    0.00000    8.00000 
        84     52    1.46154    4.69231 
        85     19    1.33771    7.00000 
        86     75    1.29656    4.16000 
        87      8    0.66144    6.75000 
        88     39    0.22057    2.05128 
        89      4    0.00000    6.00000 
        90     64    1.12500    5.87500 
        91     21    0.48562    7.95238 
        92     18    0.52411    5.94444 
        93     23    1.95604    3.00000 
        94     59    0.38724    1.05085 
        95     14    0.00000    2.00000 
        96     10    0.64031    4.30000 
        97     12    0.59512    2.25000 
        27     12    2.30940    5.00000 
        98     69    0.23902    4.02899 
        99      1    0.00000    8.00000 
        53     18    0.82589    6.61111 
       100    150    1.10079    6.96000 
       101    253    1.06823    8.31621 
       102     55    1.71488    6.49091 
       103     21    2.24669    4.00000 
        56     10    1.24900    4.80000 
       104     24    0.39965    2.08333 
       105      3    0.00000    0.00000 
        58      6    1.21335    2.83333 
       106     59    1.05385    5.64407 
       107      6    0.47140    8.33333 
      
  • English rules

  • Sequence

  • Matrix

    Target information
    Name: VAR257
    Label:
    Measurement: interval

    Tree settings

    No profile information defined
    Splitting criterion: F Test
    Significance Level: 0.2
    Minimum number of observations in a leaf: 1
    Observations required for a split search: 20
    Maximum number of branches from a node: 2
    Maximum depth of tree: 6
    Splitting rules saved in each node: 5
    Surrogate rules saved in each node: 0
    Treat missing as an acceptable value
    Model assessment measure: Average Squared Error
    Subtree: Best assessment value
    Observations sufficient for split search: 2001
    Maximum tries in an exhaustive split search: 5000
    P-value adjustment: KASS DEPTH
    Apply KASS BEFORE choosing number of branches

  • Log

  • Score Code
    Model assessment settings
    Train data set is selected for assessment.
    Validation data set is not selected for assessment.
    Test data set is not selected for assessment.
    Scored data set: 5000 observations are saved for interactive model assessment.

  • Notes: not available


    Regression

  • Parameters:

  • Fit Statistics
     
     Fit Statistic                         Training      Validation            Test 
      
     Akaike's Information Criterion    2672.1554025               .               . 
     Average Squared Error             2.9419649289               .    3.8022805964 
     Average Error Function            2.9419649289               .    3.8022805964 
     Degrees of Freedom for Error              1743               .               . 
     Model Degrees of Freedom                   257               .               . 
     Total Degrees of Freedom                  2000               .               . 
     Divisor for ASE                           2000               .            2000 
     Error Function                    5883.9298577               .    7604.5611928 
     Final Prediction Error            3.8095323261               .               . 
     Maximum Absolute Error            6.0212193739               .    6.6692721366 
     Mean Square Error                 3.3757486275               .    3.8022805964 
     Sum of Frequencies                        2000               .            2000 
     Number of Estimate Weights                 257               .               . 
     Root Average Sum of Squares       1.7152157091               .    1.9499437419 
     Root Final Prediction Error       1.9518023276               .               . 
     Root Mean Squared Error           1.8373210464               .    1.9499437419 
     Schwarz's Bayesian Criterion      4111.5873346               .               . 
     Sum of Squared Errors             5883.9298577               .    7604.5611928 
     Sum of Case Weights Times Freq            2000               .            2000 
      

  • Target Information:

    No profile information defined

  • Regression Settings:

  • Output

  • Log

  • Training Code

  • Score Code
    Model assessment settings
    Train data set is selected for assessment.
    Validation data set is not selected for assessment.
    Test data set is not selected for assessment.
    Scored data set: 5000 observations are saved for interactive model assessment.

  • Notes: not available


    Neural Network

     
     Fit Statistic                     Training    Validation        Test 
      
     [ TARGET=VAR257 ]                      .           .             . 
     Average Error                         8.25         .            8.25 
     Average Squared Error                 8.25         .            8.25 
     Sum of Squared Errors             16500.00         .        16500.00 
     Root Average Squared Error            2.87         .            2.87 
     Root Final Prediction Error           6.17         .             . 
     Root Mean Squared Error               4.81         .            2.87 
     Error Function                    16500.00         .        16500.00 
     Mean Squared Error                   23.14         .            8.25 
     Maximum Absolute Error                4.50         .            4.50 
     Final Prediction Error               38.03         .             . 
     Divisor for ASE                    2000.00         .         2000.00 
     Model Degrees of Freedom           1287.00         .             . 
     Degrees of Freedom for Error        713.00         .             . 
     Total Degrees of Freedom           2000.00         .             . 
     Sum of Frequencies                 2000.00         .         2000.00 
     Sum Case Weights * Frequencies     2000.00         .         2000.00 
     Akaike's Information Criterion     6794.43         .             . 
     Schwarz's Baysian Criterion       14002.79         .             . 
      
  • Network settings

    No profile information defined

  • Variables

  • Output

  • Log

  • Training Code

  • Score Code

    Model assessment settings
    Train data set is selected for assessment.
    Validation data set is not selected for assessment.
    Test data set is not selected for assessment.
    Scored data set: 5000 observations are saved for interactive model assessment.


    Neural Network

    Optimization plot:

    Optimization

     
     Fit Statistic                     Training    Validation       Test 
      
     [ TARGET=VAR257 ]                      .           .            . 
     Average Error                         2.95         .           3.79 
     Average Squared Error                 2.95         .           3.79 
     Sum of Squared Errors              5892.81         .        7580.65 
     Root Average Squared Error            1.72         .           1.95 
     Root Final Prediction Error           1.95         .            . 
     Root Mean Squared Error               1.84         .           1.95 
     Error Function                     5892.81         .        7580.65 
     Mean Squared Error                    3.38         .           3.79 
     Maximum Absolute Error                6.08         .           6.59 
     Final Prediction Error                3.82         .            . 
     Divisor for ASE                    2000.00         .        2000.00 
     Model Degrees of Freedom            257.00         .            . 
     Degrees of Freedom for Error       1743.00         .            . 
     Total Degrees of Freedom           2000.00         .            . 
     Sum of Frequencies                 2000.00         .        2000.00 
     Sum Case Weights * Frequencies     2000.00         .        2000.00 
     Akaike's Information Criterion     2675.17         .            . 
     Schwarz's Baysian Criterion        4114.60         .            . 
      
  • Network settings

    No profile information defined

  • Variables

  • Output

  • Log

  • Training Code

  • Score Code

    Model assessment settings
    Train data set is selected for assessment.
    Validation data set is not selected for assessment.
    Test data set is not selected for assessment.
    Scored data set: 5000 observations are saved for interactive model assessment.


    Neural Network

    Optimization plot:

    Optimization

     
     Fit Statistic                     Training    Validation       Test 
      
     [ TARGET=VAR257 ]                      .           .            . 
     Average Error                         2.37         .           4.49 
     Average Squared Error                 2.37         .           4.49 
     Sum of Squared Errors              4737.67         .        8989.83 
     Root Average Squared Error            1.54         .           2.12 
     Root Final Prediction Error           1.75         .            . 
     Root Mean Squared Error               1.65         .           2.12 
     Error Function                     4737.67         .        8989.83 
     Mean Squared Error                    2.72         .           4.49 
     Maximum Absolute Error                7.08         .           7.40 
     Final Prediction Error                3.07         .            . 
     Divisor for ASE                    2000.00         .        2000.00 
     Model Degrees of Freedom            259.00         .            . 
     Degrees of Freedom for Error       1741.00         .            . 
     Total Degrees of Freedom           2000.00         .            . 
     Sum of Frequencies                 2000.00         .        2000.00 
     Sum Case Weights * Frequencies     2000.00         .        2000.00 
     Akaike's Information Criterion     2242.80         .            . 
     Schwarz's Baysian Criterion        3693.43         .            . 
      
  • Network settings

    No profile information defined

  • Variables

  • Output

  • Log

  • Training Code

  • Score Code

    Model assessment settings
    Train data set is selected for assessment.
    Validation data set is not selected for assessment.
    Test data set is not selected for assessment.
    Scored data set: 5000 observations are saved for interactive model assessment.


    Assessment

    End Report