This content has been marked as final. Show 3 replies
Does this question relate to ODBC in some way? It looks like you posted similar questions in a number of different forums so I'm guessing that this has nothing to do with ODBC. If you haven't found the right forum yet, it would probably be helpful to take a step back and explain the basics to us-- what Oracle Corporation product are you using, for example. If you can give us more background, we can probably help you figure out what forum to post this question in.
Thanks for responding.
I'm using Oracle SQL Developer.
I'm clustering on a sample data set that came with SQL developer as a test INSUR_CUST_LTV_SAMPLE. Essentially what I want to be able to do at the end on my analysis is to create an automated re-run on an updated data set so am looking for the rules the SQL code behind the clustering is using. I hope that explains what I'm looking for a bit better?
On a side note, again I posted this questions on a few different forms but maybe you can help me. Is there of select a minimum amount of 'customers' allocated to a leaf? That sample data, when clust build is applied allocates 1 customer per leaf in places. Which doesn't really tell you anything... so is there a way of selecting a minimum as a percentage of the overall data or some rule of thumb for smallest amount per cluster?
I'm still not sure that I understand what you are doing...
You're using SQL Developer. Are you using Oracle Data Mining? Unfortunately, Oracle the corporation and Oracle the database use the term "cluster" in a lot of different places so it has a lot of potential meanings. It sounds like you're trying to run a clustering algorithm using Oracle Data Miner on a sample data set.
If that's the case, you probably want to post this over in the Data Mining