DM: DM Variable limits

In our data mining consulting practice we have frequently had to work with
databases containing over 2,000 variables and occasionally with over 5,000.
That is one reason we took great pains to ensure that our tools (CART(r)
decision trees and MARS(tm) regression) could comfortably handle over 8,000
predictor variables.

We have heard from a number of our clients that several DM suites from
major vendors are practically incapable of handling very wide files, regardless
of what their documentation states.

As an side, for those in the pharmaceutical industry, we have plans to allow
CART(r) to handle up to 1 million predictors.


  *---------------------------+---------------------------------*
  | Dan Steinberg             | FAX (619) 543 8888              |
  | Salford Systems           | VOICE (619) 543-8880            |
  | 8880 Rio San Diego Dr     |                                 |
  | San Diego, CA 92108       | http://www.salford-systems.com  |
  *-------------------------------------------------------------*

• Prev by Date: DM: Data-mining companies based in London
• Next by Date: DM: RE: Application specific mailing list: Enterprise Miner
• Prev by thread: DM: RE: Application specific mailing list: Enterprise Miner
• Next by thread: DM: Data-mining companies based in London
• Index(es):
  • Date
  • Thread