Skip to content

Latest commit

 

History

History
11 lines (6 loc) · 423 Bytes

README.md

File metadata and controls

11 lines (6 loc) · 423 Bytes

Readme

Source code for a convex surrogate operator for non-modular loss function described in:

Yu, J. and M. B. Blaschko: A Convex Surrogate Operator for General Non-Modular Loss Functions. AISTATS, 2016

An example implementation is in main.m which trains with the non-modular loss function(s) defined in testLossFunc.m

See the comments in each script for details.

If you use this code please cite the paper above.