August 15th, 2012
it is a bit silent around here. Is there any more development going on or has the project stopped?
Yup, we are still working on the project, though it has been a while since we’ve made a release. We should try to get another version out.
Do you have any questions or thoughts for future directions?
Thanks, : )
In general I really like the API. I used it for a classification/regression problem and for the statistical confidence tests in my master thesis (thanks for that ;)).
But if I could make a wish, I would find some data mining capabilities like an automated feature selection useful. But I understand that this is probably not the focus of this project.
Cool. Do you have a link or a publication to the work you’d be willing to share? Always interesting to see how people are using the Foundry.
I think feature selection is definitely within the scope of the project, we just don’t happen to have any algorithms for doing it implemented yet. What kind of algorithms were you thinking of? Top-down? Bottom up? L1 regularization? I’m know there are lots of ways to do feature selection, I’m curious what the most commonly used ones are.
We have been busy added a whole lot of functionality to the Foundry, including a lot of additional algorithms, and a new way to allow direct community contribution to the Foundry code base. We hope to have the release cut soon!
Let us know if anybody has any requests.
You must be logged in to reply to this topic.