Dear Sir/Madam,
I would like to understand - What makes the outcomes different when we use
1. <SAS Enterprise Miner 14.1> 'Decision Tree'
2. <SAS Enterprise Miner 14.1> 'High-Performance Tree'
3. <SAS 14> Tree with the HPSPLIT Procedure
I hope the developer/someone (knows the algorithm) explain the reasons why the outcomes (including misclassification rates) are different, even with the same options and the data.
a. I understood 2 is using the HPSPLIT Procedure in SAS Enterprise Miner.
Even with the same options and the data, the outcomes from 2 and 3 are different.
b.It seems SAS Enterprise Miner suggests (slighly) lower misclassification rates than the outcomes from SAS 14 Tree models - I am wondering SAS Enterprise Miner recommanded to use for tree models.
Thank you very much, in advance.
