SAS Institute Alert Note SN-007141

From: SAS Technical Support (LISTSERV@VM.SAS.COM)
Date: Thu Apr 11 2002 - 10:57:12 EDT

  • Next message: SAS Technical Support: "SAS Institute TS: New Hot Fix Releases"

    SN-007141 ***Alert Note***

    Scores from neural networks incorrect when the value of the frequency
    variable is nonpositive or missing

      Product: Enterprise Miner
    Component: Neural Network node
     Priority: ALERT

      In the Neural Network node of Enterprise Miner, the values of new
      variables added from the "Process or Score Data" option in the Output
      tab, such as predicted values and their associated residuals and error
      function values, will be incorrect for observations with nonpositive or
      missing frequency values. Because these observations are not used in
      the analysis, the model fit statistics and parameters are correct.

      In order to score these observations the scoring must take place outside
      of the Neural Network node. First extract these observations into a
      separate data set using the Sampling or SAS Code node, and then score
      them using the Score node.

    Keywords:
      neural network score out predicted residual error function incorrect
      alert frequency freq missing zero output process incorrout incorrstat
      scores from neural networks incorrect when the value of the frequency
      variable is nonpositive or missing
      007141 7141

    SAS Note Revised On: Wed, 20 Mar 2002

    System Version Reported Version Fixed

    AIX/6000 4.0
    AIX/6000 4.1
    Compaq Tru64 UNIX 4.0
    Compaq Tru64 UNIX 4.1
    HP-UX Operating Systems 4.0
    HP-UX Operating Systems 4.1
    Solaris 4.0
    Solaris 4.1
    Windows NT 4.0
    Windows NT 4.1
    Windows 95 4.0
    Windows 95 4.1

    No Fixes Available

    +-------------------------------------------------------------------------
    + NOTE: To unsubscribe you can reply to this mail with:
    + "SIGNOFF tsnews-l"
    + as the only text in the body of the message (without the double quotes).
    + If you have any questions please send them to SUPPORT@SAS.COM
    +-------------------------------------------------------------------------



    This archive was generated by hypermail 2b29 : Thu Apr 11 2002 - 10:58:18 EDT