Orange Forum • View topic - kernel ridge regression and penalized logistic regression

kernel ridge regression and penalized logistic regression

A place to ask questions about methods in Orange and how they are used and other general support.

kernel ridge regression and penalized logistic regression

Postby psederberg » Thu Sep 27, 2007 16:03

Hi Everybody:

I saw a similar post on the wishlist forum, but perhaps this is the better forum to ask this on.

Is there any code in orange for running kernel ridge regression and penalized logistic regression. I've used the logistic regression code with no problems, but I can't find these other regressions in orange.

If they are not in orange, could someone point me in the right direction for how to implement them in orange C++ code so that they can run quickly (I know how to make a python learner from the tutorials, but it's way to slow.)

Finally, I'd like to be able to use a pre-computed custom kernel matrix in the kernel methods already implemented (such as svm), is there any existing support for this or do I need to implement a custom kernel that performs a lookup into my precomputed kernel?

Thanks for any help,
Per

Postby Guest » Sat Oct 06, 2007 19:14

Is anybody working on this or can provide pointers on how to best implement this functionality?

Thanks for any help!


Return to Questions & Support