# Predicting Partial Orders: Ranking with Abstention

@inproceedings{Cheng2010PredictingPO, title={Predicting Partial Orders: Ranking with Abstention}, author={Weiwei Cheng and Micha{\"e}l Rademaker and Bernard De Baets and Eyke H{\"u}llermeier}, booktitle={ECML/PKDD}, year={2010} }

The prediction of structured outputs in general and rankings in particular has attracted considerable attention in machine learning in recent years, and different types of ranking problems have already been studied. In this paper, we propose a generalization or, say, relaxation of the standard setting, allowing a model to make predictions in the form of partial instead of total orders. We interpret such kind of prediction as a ranking with partial abstention: If the model is not sufficiently… Expand

#### 54 Citations

Label Ranking with Abstention: Predicting Partial Orders by Thresholding Probability Distributions (Extended Abstract)

- Computer Science
- ArXiv
- 2011

A new method for learning to predict partial orders that improves on an existing approach, both theoretically and empirically, is proposed, based on the idea of thresholding the probabilities of pairwise preferences between labels as induced by a predicted (parameterized) probability distribution on the set of all rankings. Expand

Label Ranking with Partial Abstention based on Thresholded Probabilistic Models

- Computer Science, Mathematics
- NIPS
- 2012

This work addresses abstention for the label ranking setting, allowing the learner to declare certain pairs of labels as being incomparable and, thus, to predict partial instead of total orders. Expand

A Margin-based MLE for Crowdsourced Partial Ranking

- Computer Science, Mathematics
- ACM Multimedia
- 2018

This paper proposes a novel framework to learn some probabilistic models of partial orders as a margin-based Maximum Likelihood Estimate (MLE) method, and proves that the induced MLE is a joint convex optimization problem with respect to all the parameters, including the global ranking scores and margin parameter. Expand

Cautious Label-Wise Ranking with Constraint Satisfaction

- Computer Science
- IPMU
- 2020

This paper proposes to combine a rank-wise decomposition, in which every sub-problem becomes an ordinal classification one, with a constraint satisfaction problem (CSP) approach to verify the consistency of the predictions. Expand

iSplit LBI: Individualized Partial Ranking with Ties via Split LBI

- Computer Science, Mathematics
- NeurIPS
- 2019

This paper formulate a unified framework which not only can be used for individualized partial ranking prediction, but can also be helpful for abnormal users selection, which is realized by a variable splitting-based algorithm called iSplit LBI. Expand

Modelling and predicting partial orders from pairwise belief functions

- Computer Science, Mathematics
- Soft Comput.
- 2016

In this paper, we introduce a generic way to represent and manipulate pairwise information about partial orders (representing rankings, preferences, ...) with belief functions. We provide generic and… Expand

A Pairwise Label Ranking Method with Imprecise Scores and Partial Predictions

- Computer Science, Mathematics
- ECML/PKDD
- 2013

This paper proposes a ranking method based on pairwise imprecise scores obtained from likelihood functions, which can be aggregated to produce interval orders, which are specific types of partial orders. Expand

Multilabel predictions with sets of probabilities: The Hamming and ranking loss cases

- Mathematics, Computer Science
- Pattern Recognit.
- 2015

It is shown that when considering the Hamming or the ranking loss, outer-approximating predictions can be efficiently computed from label-wise information, as in the precise case. Expand

Dyad ranking using Plackett–Luce models based on joint feature representations

- Computer Science
- Machine Learning
- 2017

This paper proposes an extension of an existing label ranking method based on the Plackett–Luce model, a statistical model for rank data, combined with a suitable feature representation of dyads that allows for learning a (highly nonlinear) joint feature representation. Expand

Lexicographic preferences for predictive modeling of human decision making: A new machine learning method with an application in accounting

- Mathematics, Computer Science
- Eur. J. Oper. Res.
- 2017

This paper introduces a learning algorithm for inducing generalized lexicographic preference models from a given set of training data, which consists of pairwise comparisons between objects, and generalizes simpleLexicographic orders in the sense of allowing the model to consider several attributes simultaneously (instead of looking at them one by one), thereby significantly increasing the expressiveness of the model class. Expand

#### References

SHOWING 1-10 OF 21 REFERENCES

Pairwise Preference Learning and Ranking

- Computer Science
- ECML
- 2003

The main objective of this work is to investigate the trade-off between the quality of the induced ranking function and the computational complexity of the algorithm, both depending on the amount of preference information given for each example. Expand

Learning to Order Things

- Computer Science, Mathematics
- NIPS
- 1997

An on-line algorithm for learning preference functions that is based on Freund and Schapire's "Hedge" algorithm is considered, and it is shown that the problem of finding the ordering that agrees best with a learned preference function is NP-complete. Expand

Binary Decomposition Methods for Multipartite Ranking

- Mathematics, Computer Science
- ECML/PKDD
- 2009

This paper discusses extensions of the AUC metric which are suitable as evaluation criteria for multipartite rankings and proposes methods on the basis of binary decomposition techniques that have previously been used for multi-class and ordinal classification. Expand

Label ranking by learning pairwise preferences

- Mathematics, Computer Science
- Artif. Intell.
- 2008

This work shows that a simple (weighted) voting strategy minimizes risk with respect to the well-known Spearman rank correlation and compares RPC to existing label ranking methods, which are based on scoring individual labels instead of comparing pairs of labels. Expand

Preference Learning

- Computer Science
- Künstliche Intell.
- 2005

This article aims at conveying a first idea of typical preference learning problems, namely learning from label preferences and learning from object preferences. Expand

Preference Learning

- Computer Science
- 2010

The editors first offer a thorough introduction, including a systematic categorization according to learning task and learning technique, along with a unified notation, and the first half of the book is organized into parts on applications of preference learning in multiattribute domains, information retrieval, and recommender systems. Expand

Log-Linear Models for Label Ranking

- Computer Science, Mathematics
- NIPS
- 2003

This work presents a general boosting-based learning algorithm for the label ranking problem and proves a lower bound on the progress of each boosting iteration. Expand

A threshold for majority in the context of aggregating partial order relations

- Mathematics, Computer Science
- International Conference on Fuzzy Systems
- 2010

A voting problem where voters have expressed their preferences on a single set of objects takes the shape of strict partial order relations, and the minimum number of votes a pairwise preference should receive in order to qualify as a social pairwise preferences is determined. Expand

Classification with a Reject Option using a Hinge Loss

- Computer Science, Mathematics
- J. Mach. Learn. Res.
- 2008

This work considers the problem of binary classification where the classifier can, for a particular cost, choose not to classify an observation and proposes a certain convex loss function φ, analogous to the hinge loss used in support vector machines (SVMs). Expand

Classification with reject option

- Mathematics
- 2006

This paper studies two-class (or binary) classification of elements X in R k that allows for a reject option. Based on n independent copies of the pair of random variables (X,Y ) with X 2 R k and Y 2… Expand