Consistent Robust Adversarial Prediction for General Multiclass Classification

Rizal Fathony
Rizal Fathony
Kaiser Asif
Kaiser Asif
Mohammad Ali Bashiri
Mohammad Ali Bashiri
Sima Behpour
Sima Behpour

arXiv: Machine Learning, Volume abs/1812.07526, 2018.

Cited by: 4|Views56
EI

Abstract:

We propose a robust adversarial prediction framework for general multiclass classification. Our method seeks predictive distributions that robustly optimize non-convex and non-continuous multiclass loss metrics against the worst-case conditional label distributions (the adversarial distributions) that (approximately) match the statistics ...More

Code:

Data:

Your rating :
0

 

Tags
Comments