Bayes-Optimal Scorers for Bipartite Ranking
نویسندگان
چکیده
We address the following seemingly simple question: what is the Bayes-optimal scorer for a bipartite ranking risk? The answer to this question helps elucidate the relationship between bipartite ranking and other established learning problems. We show that the answer is non-trivial in general, but may be easily determined for certain special cases using the theory of proper losses. Our analysis immediately establishes equivalence relationships between several seemingly disparate approaches to bipartite ranking, such as minimising a suitable class-probability estimation risk, and minimising the p-norm push risk proposed in Rudin (2009).
منابع مشابه
Bipartite Ranking: a Risk-Theoretic Perspective
We present a systematic study of the bipartite ranking problem, with the aim of explicating its connections to the class-probability estimation problem. Our study focuses on the properties of the statistical risk for bipartite ranking with general losses, which is closely related to a generalised notion of the area under the ROC curve: we establish alternate representations of this risk, relate...
متن کاملBipartite ranking: risk, optimality, and equivalences
We present a systematic study of the bipartite ranking problem, with the aim of delineating its connections to the class-probability estimation problem. Our study focuses on the properties of the statistical risk for bipartite ranking, which is closely related to the area under the ROC curve: we establish alternate representations of the risk, relate the Bayes-optimal risk to a class of probabi...
متن کاملOn Theoretically Optimal Ranking Functions in Bipartite Ranking
This paper investigates the theoretical relation between loss criteria and the optimal ranking functions driven by the criteria in bipartite ranking. In particular, the relation between AUC maximization and minimization of ranking risk under a convex loss is examined. We characterize general conditions for ranking-calibrated loss functions in a pairwise approach, and show that the best ranking ...
متن کاملUpper bounds and aggregation in bipartite ranking
One main focus of learning theory is to find optimal rates of convergence. In classification, it is possible to obtain optimal fast rates (faster than n−1/2) in a minimax sense. Moreover, using an aggregation procedure, the algorithms are adaptive to the parameters of the class of distributions. Here, we investigate this issue in the bipartite ranking framework. We design a ranking rule by aggr...
متن کاملOptimal edge ranking of complete bipartite graphs in polynomial time
An edge ranking of a graph is a labeling of edges using positive integers such that all paths connecting two edges with the same label visit an intermediate edge with a higher label. An edge ranking of a graph is optimal if the number of labels used is minimum among all edge rankings. As the problem of finding optimal edge rankings for general graphs is NP-hard [12], it is interesting to concen...
متن کامل