On Saturday, February 25, 2006, 1:38:53 PM, Joe wrote: JW> JW> JW> I would actually prefer that MDLP autotune the weight for JW> invURIBL, but since the weights are managed by invURIBL and not JW> Declude I don't know how this will work.
I'm not familiar enough with invURIBL to know how it is configured. However, as long as it's maximum and minumum weights are in a reasonable range, then if you exclude it from MDLP you should be ok. MDLP's AI tries to optimize the weights of the tests it can manipulate so that the most accurate total scores are provided. If there are tests it cannot adjust then it is forced to work around those with the other tests. The results are not predictable (the task is far too dynamic and contains far too many variables) but they should be sane and correct. _M This E-Mail came from the Message Sniffer mailing list. For information and (un)subscription instructions go to http://www.sortmonster.com/MessageSniffer/Help/Help.html
