Bipartite Ranking through Minimization of Univariate Loss

By W. Kotłowski, Krzysztof Dembczynski, Eyke Hüllermeier. 2011

100 citations recorded by the source corpus.

Subfields: Information Retrieval & Search, Statistical Learning Theory

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Information Retrieval & Search

In Statistical Learning Theory

Citations in the corpus

Learning Optimized Risk Scores (2016, background): “[61] Kotlowski, Wojciech, Krzysztof J Dembczynski, and Eyke Huellermeier.”
Convex Calibration Dimension for Multiclass Loss Matrices (2014, background): “…in understanding consistency properties of surrogate risk minimization algorithms for various learning problems, such as binary clas-siﬁcation, multiclass 0-1 classiﬁcation, and various forms of ranking and multi-label prediction problems [18, 14, 34, 26, 2, 33, 28, 27, 8, 31, 10, 23, 4, 12, 15].”
"On the (Non-)existence of Convex, Calibrated Surrogate Losses for Ranking" (2012, methodology): “The calibration of surrogate losses has been extensively studied for various classification settings [1, 26, 27, 18, 19] and for AUC optimization [7, 15].”
Direct Learning to Rank and Rerank (2018, background): “The example considered in this section also illustrates why arguments about consistency (or lack thereof) of ranking methods (e.g., Kotlowski et al, 2011) are not generally relevant for practice.”
Optimized Risk Scores (2017, background): “[20] show that a “balanced” logistic loss is a lower bound on 1-AUC, which means that minimizing the logistic loss maximizes a surrogate of AUC.”
Predicting accurate probabilities with a ranking loss (2012, background): “ut in this case there is no clear reason to eschew kernel logistic regression. Predicting accurate probabilities with a ranking loss The issue of how best to maximize AUC is not settled. For example, Kotlowski et al. (2011) show that the ranking error (viz. 1 A ) of a model can be upper bounded by i”
Surrogate regret bounds for bipartite ranking via strongly proper losses (2012, methodology): “Our proof technique is considerably simpler than that of [19], and relies on properties of proper (composite) losses as elucidated recently for example in [22, 23, 15, 6].”
A Review on Multi-Label Learning Algorithms (2014, methodology): “By using a reduction to the bipartite ranking problem [55], simple univariate convex surrogate loss (exponential or logistic) defined on single labels is shown to be consistent with the ranking loss with explicit regret bounds and convergence rates.”