Welcome to the technical support knowledge base for ADAPA on the Cloud. Our blogs cover general questions and information related to predictive models, PMML, and supported functionality of the ADAPA decision engine. Please use the search tool or the FAQ Categories to the left to find the information you are looking for. If you can't find it, feel free to contact us.

© Zementis, Inc. - All Rights Reserved.

Friday, June 5, 2009

Problems with KXEN PMML exported models.


The latest KXEN software exports perfect PMML, however if you are using older versions of KXEN, the PMML model it exports may have some problems which will be automatically fixed by the PMML Converter during conversion. So, you should convert your model (even if it is already expressed in the latest version of PMML) before uploading it in ADAPA.

If you have a Mining Model (easy to locate in the PMML file - look for the MiningModel element) however, check our blog on how to upload KXEN Mining Models into ADAPA.

Here is a list of issues we encountered while evaluating old KXEN PMML models (together with the remedies):
  1. Your PMML model needs to contain the URL with the address of the PMML schema. Our PMML Converter will add that to the model automatically once you pass it through the converter.

  2. Models may contain DerivedFields for which optype = "continuous" but dataType = "string". Just change the dataType to "double".

  3. Models may contain DerivedFields in which the output of a NormContinuous transformation is a float (dataType = "float"). Change the dataType to "double".

  4. For clustering models, make sure compareFunction = "absdiff" is expressed with a small "d". Models may refer to "absDiff" instead which is not valid.

  5. Again, for clustering models only, delete the element CenterFields (this is not valid PMML).
After all the remedies have been applied (feel free to use the PMML Converter for that), your model should be ready for uploading into ADAPA.

0 comments:






Copyright © 2009 Zementis Incorporated. All rights reserved.

Privacy - Terms Of Use - Contact Us