In a rather complex scoring stream I am using an XGBoosted Tree, Boosted C&RT, C5.0, Logistic, and CHAID models to predict a rare event. For an unknown reason, the XGBT model scores the first five records out of over 4000 and that's it. No error message is produced. Other Model Applier nodes branching from the same Type node work as expected. There is no sampling and no subsetting going on in the stream. What could be causing this?
This is on windows 10 and version 18.1.1
Modeler licenses from more vendors 1 Answer
How to check the maximum logging number of SPSS in use from License Manager usage file? 1 Answer
Duplicated commuter licenses after server reboot 1 Answer
How does SPSS concurrent licensing work when using VMware horizon instant clones? 1 Answer
Change license laptop. 1 Answer