On Generalizing the C-Bound to the Multiclass and Multi-label Settings

01/13/2015

∙

The C-bound, introduced in Lacasse et al., gives a tight upper bound on the risk of a binary majority vote classifier. In this work, we present a first step towards extending this work to more complex outputs, by providing generalizations of the C-bound to the multiclass and multi-label settings.

READ FULL TEXT

On Generalizing the C-Bound to the Multiclass and Multi-label Settings

Understanding Label Bias in Single Positive Multi-Label Learning

Classification of sparse binary vectors

Convex Calibrated Surrogates for the Multi-Label F-Measure

On a scalable problem transformation method for multi-label learning

Characterizing the Evasion Attackability of Multi-label Classifiers

Multi-winner Approval Voting Goes Epistemic

Attack Transferability Characterization for Adversarially Robust Multi-label Classification

On Generalizing the C-Bound to the Multiclass and Multi-label Settings

Related Research

Understanding Label Bias in Single Positive Multi-Label Learning

Classification of sparse binary vectors

Convex Calibrated Surrogates for the Multi-Label F-Measure

On a scalable problem transformation method for multi-label learning

Characterizing the Evasion Attackability of Multi-label Classifiers

Multi-winner Approval Voting Goes Epistemic

Attack Transferability Characterization for Adversarially Robust Multi-label Classification