How can I use Hinge loss for Multi-class Support Vector Machine algorithm?

I have coded a Multi-class SVM classifier using Hinge loss as loss function. I worked on the knowledge in this article: https://cs231n.github.io/linear-classify/.
But instead of decreasing and converging, my loss function increases to infinity after few loops.
I attach here file .m. Could anyone show me how I can fix this bug?

Respuestas (0)

Categorías

Más información sobre Statistics and Machine Learning Toolbox en Centro de ayuda y File Exchange.

Preguntada:

el 7 de Mayo de 2018

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by