Hi,
Does genie have a way of estimating the influence of each training sample on the model? or, a method for finding the most influential training samples? (for example, in the sense that without them prediction will change the most, or model's parameters will change the most).
Thank you.
Influential samples
-
- Site Admin
- Posts: 430
- Joined: Tue Dec 11, 2007 4:24 pm
Re: Influential samples
No, I'm afraid there is nothing like that in GeNIe. The idea is interesting, though!