Support Vector Machine: insert self-made kernel (Feature #649)


Added by Walter Lucetti over 6 years ago. Updated over 1 year ago.


Status:Open Start date:
Priority:Normal Due date:
Assignee:- % Done:

0%

Category:ml
Target version:Next Hackathon
Difficulty: Pull request:

Description

It would be useful to be able to insert self-made kernel in SVM structure.

Actual implementation provides only RBF, Sigmoid and polynomial kernels in SVM, but this is too restrictive since in literature you can find a lot of new kernels studied to solve some kind of problems.

You could insert a "USER-KERNEL" option in available kernel and a callback prototipe to implement user-kernel.

Walter


History

Updated by Alexander Smorkalov about 3 years ago

  • Priority changed from High to Normal
  • Target version set to Next Hackathon

Updated by Maksim Shabunin over 1 year ago

Issue has been transferred to GitHub: https://github.com/Itseez/opencv/issues/4255

Also available in: Atom PDF