Advertisement
Published in:
2013 | OriginalPaper | Chapter
19. Fast Training for Multi-Class Classification: K-SVCR
Authors : Zhibin Zhu, Anwa Zhou
Published in: Informatics and Management Science IV
Publisher: Springer London
Activate our intelligent search to find suitable subject content or patents.
Select sections of text to find matching patents with Artificial Intelligence. powered by
Select sections of text to find additional relevant content using AI-assisted search. powered by
Abstract
In this paper, we apply a regularized no smooth Newton method to solve the multi-class classification \( K{\text{-}}SVCR \) machine which is reformulated as an box constrained variation inequality problem (BVIP) with a positive semi-definite matrix. This algorithm fully exploits the typical feature of sparsity solution for the \( K{\text{-}}SVCR \) method, which shows that our algorithm for \( K{\text{-}}SVCR \) can be implemented efficiently. Preliminary numerical results on benchmark data sets indicate that the method is remarkably faster than the standard Mat lab routine for training the original \( K{\text{-}}SVCR \).