Completing Preferences by Means of Analogical Proportions

We suppose that all we know about the preferences of an agent, is given by a (small) collection of relative preferences between choices represented by their evaluations on a set of criteria. Taking lesson from the success of the use of analogical proportions for predicting the class of a new item from a set of classified examples, we explore the possibility of using analogical proportions for completing a set of relative preferences. Such an approach is also motivated by a striking similarity between the formal structure of the axiomatic characterization of weighted averages and the logical definition of an analogical proportion. This paper discusses how to apply an analogical proportion-based approach to the learning of relative preferences, assuming that the preferences are representable by a weighted average, and how to validate experimental results. The approach is illustrated by examples.

[1]  Didier Dubois,et al.  Expressing Preferences from Generic Rules and Examples - A Possibilistic Approach Without Aggregation Function , 2005, ECSQARU.

[2]  Salvatore Greco,et al.  Ordinal regression revisited: Multiple criteria ranking using a set of additive value functions , 2008, Eur. J. Oper. Res..

[3]  Henri Prade,et al.  Handling Analogical Proportions in Classical Logic and Fuzzy Logics Settings , 2009, ECSQARU.

[4]  Mathieu Serrurier,et al.  Elicitating Sugeno Integrals: Methodology and a Case Study , 2009, ECSQARU.

[5]  François Yvon,et al.  Formal Models of Analogical Proportions , 2007 .

[6]  Gilles Richard,et al.  Analogical classification: A new way to deal with examples , 2014, ECAI.

[7]  Eyke Hüllermeier,et al.  Preference Learning (Dagstuhl Seminar 14101) , 2014, Dagstuhl Reports.

[8]  Gilles Richard,et al.  Reasoning with Logical Proportions , 2010, KR.

[9]  Henri Prade,et al.  Ranking Alternatives on the Basis of Generic Constraints and Examples - A Possibilistic Approach , 2007, IJCAI.

[10]  Peter P. Wakker,et al.  Additive Representations of Preferences , 1989 .

[11]  H. Prade,et al.  A Choquet Integral Representation in Multicriteria Decision Making , 2002 .

[12]  Laurent Miclet,et al.  Learning by Analogy: A Classification Rule for Binary and Nominal Data , 2007, IJCAI.

[13]  Matthieu Cord,et al.  Quadruplet-Wise Image Similarity Learning , 2013, 2013 IEEE International Conference on Computer Vision.

[14]  Michel Grabisch,et al.  A review of methods for capacity identification in Choquet integral based multi-attribute utility theory: Applications of the Kappalab R package , 2008, Eur. J. Oper. Res..

[15]  Jean-Yves Béziau.  From analogical proportion to logical proportions , 2013 .

[16]  Yannis Siskos,et al.  Preference disaggregation: 20 years of MCDA experience , 2001, Eur. J. Oper. Res..

[17]  P. Wakker.  Additive Representations of Preferences: A New Foundation of Decision Analysis , 1988 .

[18]  J. Siskos.  Assessing a set of additive utility functions for multicriteria decision-making , 1982 .

[19]  Gilles Richard,et al.  Multiple-valued extensions of analogical proportions , 2016, Fuzzy Sets Syst..