Skip to Main Content
Forums

Data Science & Machine Learning

Announcement

For appeals, questions and feedback about Oracle Forums, please email oracle-forums-moderators_us@oracle.com. Technical questions should be asked in the appropriate category. Thank you!

storage from swl dev

981516Apr 21 2013 — edited Apr 24 2013
hi
how I can storage model result such as clustering to special table that I use it's data for clustering? I work with SQL DEL. oracle

Edited by: Nasiri Mahdi on Apr 20, 2013 9:20 PM
This post has been answered by Mark Kelly-Oracle on Apr 21 2013
Jump to Answer

Comments

Mark Kelly-Oracle
Answer
Hi Nasiri,
If you want to score(apply) data using a Cluster model, then you should use the Apply Node.
Connect the Cluster Model node and a data flow node to the Apply node and select whatever type of prediction result you want.
If you simply want to store model details from a Cluster model then connect the cluster model node to a Model Details node and select the desired model details type. You can then connect that result to a Create Table node
Thanks,Mark
Marked as Answer by 981516 · Sep 27 2020
981516
I connect reply node to clustering. clustering is run but reply's run is disable. can you please tell me what is my problem?
Mark Kelly-Oracle
Hi Nasiri,
Not sure if I understand the details of your problem. But it sounds like when you say "reply node" you mean "Apply Node".
An Apply node requires a model node (either a build node or a model node) as input along with a data flow input.
Perhaps you only have the model as input?
Are you trying to export model details or do a apply?
Thanks, Mark
1 person found this helpful
981516
Hi Mark
tanks for your help
I can see result via your guide. I have another question, I want to save my column of my input table and it's cluster name for any record.
another problem: I set number of cluster =4 but it give me 7 cluster ID that are hierarchical but I need cluster ID=1,2,3, or 4
Mark Kelly-Oracle
Hi Nasiri,
The ODM scoring functions will only return scores for the leaf clusters.
It is possible to sum the probabilities of the leaves that fall under a particular non-leaf cluster. Knowing the hierarchy and the cluster set probabilities, it is possible to write a query to achieve that.
What is your use case for wanting to produce such a result?
Also, in the future, please open another thread when you change topics to assist us and others in reviewing and responding to postings.
Thanks,Mark
1 - 5
Locked Post
New comments cannot be posted to this locked post.

Post Details

Locked on May 22 2013
Added on Apr 21 2013
#machine-learning
5 comments
2,564 views