PMML is the discussion topic of the next Chicago Machine Learning Meetup in a couple of days (Sept. 14).  Since there will be quite a few people there who have no prior experience with PMML, I'm wondering if there's a good 15-20 minute Powerpoint overview out there that I could borrow and present to get people up to speed. See More

The Data Dictionary for a PMML model requires quite a bit of metadata for each field.  With sparse, high dimensional data the Data Dictionary could be many times larger than either the training data or the trained model.   Has anyone developed a standard extension to the PMML syntax that, for instance, just says that all fields have the same metadata?

The optional MiningBuildTask element in PMML, which allows free-form content, seems like a great idea to me.  I often want to record information for future reference about the parameter settings used in training and potentially statistics about the training process.  I have my own thoughts about what to include in this element, but would be interested in pointers to best practices, or good concrete examples, of using this element.  Surprisingly, this element is not even discussed in the…See More

Does anyone know of examples of wrapping an existing piece of supervised learning software to output models in PMML format?  Of particular interest are learners that just take in labeled vectors of numbers as training data and put out models that are pretty much just coefficient vectors (liblinear, SVMlight, BXRtrain, BOW, etc.).  That is, they don't have any smarts about data types, ranges of legal values of features, etc.: something else is assumed to deal with that and present the learner…See More