The Structured Abstain Problem and the Lovász Hinge

03/16/2022
by   Jessie Finocchiaro, et al.
0

The Lovász hinge is a convex surrogate recently proposed for structured binary classification, in which k binary predictions are made simultaneously and the error is judged by a submodular set function. Despite its wide usage in image segmentation and related problems, its consistency has remained open. We resolve this open question, showing that the Lovász hinge is inconsistent for its desired target unless the set function is modular. Leveraging a recent embedding framework, we instead derive the target loss for which the Lovász hinge is consistent. This target, which we call the structured abstain problem, allows one to abstain on any subset of the k predictions. We derive two link functions, each of which are consistent for all submodular set functions simultaneously.

READ FULL TEXT
research
12/24/2015

The Lovász Hinge: A Novel Convex Surrogate for Submodular Losses

Learning with non-modular losses is an important problem when sets of pr...
research
07/17/2019

An Embedding Framework for Consistent Polyhedral Surrogates

We formalize and study the natural approach of designing convex surrogat...
research
07/18/2022

Consistent Polyhedral Surrogates for Top-k Classification and Variants

Top-k classification is a generalization of multiclass classification us...
research
06/29/2022

An Embedding Framework for the Design and Analysis of Consistent Polyhedral Surrogates

We formalize and study the natural approach of designing convex surrogat...
research
04/12/2016

A Convex Surrogate Operator for General Non-Modular Loss Functions

Empirical risk minimization frequently employs convex surrogates to unde...
research
11/12/2008

The Expressive Power of Binary Submodular Functions

It has previously been an open problem whether all Boolean submodular fu...
research
02/05/2019

A General Theory for Structured Prediction with Smooth Convex Surrogates

In this work we provide a theoretical framework for structured predictio...

Please sign up or login with your details

Forgot password? Click here to reset