Is PROC SPLIT the best option to create a decision tree in SAS?

[Bond007]

I want to create a decision tree in SAS. I have found that I can use PROC SPLIT, but there are other procedures such as PROC HPSPLIT, PROC ARBOR etc. 

The document (link below) - page 8, says that PROC SPLIT procedure of  Enterprise Miner is not supported by SAS.

https://pdfs.semanticscholar.org/8cb2/a1d2f3eee6ca610224371e4a7f5a608ef43f.pdf

So I'm curious about whether I should use PROC SPLIT or some other procedure for decision tree.

Thanks in advance.

Regards,

Bond

[Ksharp]

Try PROC HPSPLIT ,
But I would recommend to use Random Forest via PROC HPFOREST .

[Bond007]

I have researched about PROC SPLIT, PROC ARBOR and PROC HPSPLIT and my findings are as follows:

PROC ARBOR was introduced in SAS 9. It is mentioned in SAS documentation that it will eventually replace PROC SPLIT, as it is faster than PROC SPLIT on larger datasets.

PROC HPSPLIT was introduced in SAS 9.4

I'm trying to find differences between PROC ARBOR and PROC HPSPLIT. But I couldn't find anything concrete in SAS documentation.

Any help would be appreciated.

Thanks,

Bond

[Ksharp]

I never heard about PROC ARBOR .
The most new PROC I think is PROC HPSPLIT ,you can use it under SAS9.4 ,no need SAS/EM .