Re: Training Decision Tree
- From: "Dejan Sarka" <dejan_please_reply_to_newsgroups.sarka@xxxxxxxxxx>
- Date: Fri, 23 Mar 2007 11:08:44 +0100
But if I want SQL Server to take a specific amount of data for validation,
is there a way for that?
I am not sure if I understand what you mean. In all examples from my
previous post you can easily specify any amount you need in percentages;
with SSIS Row Sampling transform, you can also specify number of rows you
need. If you want to get a fixed number of rows in T-SQL, you can use TOP
operator, like
SELECT TOP 1000 *
FROM table
ORDER BY NEWID()
By using NEWID() function for ordering you get approximately random sample,
usually good enough for analyses; however, SSIS Row Sampling transform gives
you statisticaly random rows.
--
Dejan Sarka
http://www.solidqualitylearning.com/blogs/
.
- Follow-Ups:
- Training Decision Tree
- From: anonymous_user@xxxxxxxxxxxxxxxxxxxxxxx
- Training Decision Tree
- From: anonymous_user@xxxxxxxxxxxxxxxxxxxxxxx
- Training Decision Tree
- From: anonymous_user@xxxxxxxxxxxxxxxxxxxxxxx
- Training Decision Tree
- References:
- Training Decision Tree
- From: anonymous_user@xxxxxxxxxxxxxxxxxxxxxxx
- Re: Training Decision Tree
- From: Dejan Sarka
- Re: Training Decision Tree
- From: anonymous_user@xxxxxxxxxxxxxxxxxxxxxxx
- Training Decision Tree
- Prev by Date: Re: Training Decision Tree
- Next by Date: Where can I get the source code for XMLA thin miner??
- Previous by thread: Re: Training Decision Tree
- Next by thread: Training Decision Tree
- Index(es):