Learning Choice Functions with Gaussian Processes

In consumer theory, ranking available objects by means of preference relations yields the most common description of individual choices. However, preference-based models assume that individuals: (1) give their preferences only between pairs of objects; (2) are always able to pick the best preferred object. In many situations, they may be instead choosing out of a set with more than two elements and, because of lack of information and/or incomparability (objects with contradictory characteristics), they may not able to select a single most preferred object. To address these situations, we need a choice-model which allows an individual to express a set-valued choice. Choice functions provide such a mathematical framework. We propose a Gaussian Process model to learn choice functions from choice-data. The proposed model assumes a multiple utility representation of a choice function based on the concept of Pareto rationalization, and derives a strategy to learn both the number and the values of these latent multiple utilities. Simulation experiments demonstrate that the proposed model outperforms the state-of-the-art methods.

[1]  Marco Zaffalon,et al.  Correlated product of experts for sparse Gaussian process regression , 2021, Machine Learning.

[2]  Alessio Benavoli,et al.  A unified framework for closed-form nonparametric regression, classification, preference and mixed problems with Skew Gaussian Processes , 2020, Machine Learning.

[3]  Alessio Benavoli,et al.  Preferential Bayesian optimisation with skew gaussian processes , 2020, GECCO Companion.

[4]  Eyke Hüllermeier,et al.  Learning Choice Functions via Pareto-Embeddings , 2020, KI.

[5]  Siu Lun Chau,et al.  Learning Inconsistent Preferences with Gaussian Processes , 2020, AISTATS.

[6]  Alessio Benavoli,et al.  Skew Gaussian processes for classification , 2020, Machine Learning.

[7]  Aki Vehtari,et al.  Preferential Batch Bayesian Optimization , 2020, 2021 IEEE 31st International Workshop on Machine Learning for Signal Processing (MLSP).

[8]  Marco Zaffalon,et al.  Recursive Estimation for Sparse Gaussian Process Regression , 2019, Autom..

[9]  Gert de Cooman,et al.  Interpreting, Axiomatising and Representing Coherent Choice Functions in Terms of Desirability , 2019, ISIPTA.

[10]  Pritha Gupta,et al.  Learning context-dependent choice functions , 2019, Int. J. Approx. Reason..

[11]  Marco Zaffalon,et al.  Statistical comparison of classifiers through Bayesian hierarchical modelling , 2016, Machine Learning.

[12]  Carl E. Rasmussen,et al.  Understanding Probabilistic Sparse Gaussian Process Approximations , 2016, NIPS.

[13]  Aki Vehtari,et al.  Practical Bayesian model evaluation using leave-one-out cross-validation and WAIC , 2015, Statistics and Computing.

[14]  Daniel Hernández-Lobato,et al.  Scalable Gaussian Process Classification via Expectation Propagation , 2015, AISTATS.

[15]  James Hensman,et al.  Scalable Variational Gaussian Process Classification , 2014, AISTATS.

[16]  Neil D. Lawrence,et al.  Gaussian Processes for Big Data , 2013, UAI.

[17]  Zoubin Ghahramani,et al.  Bayesian Active Learning for Classification and Preference Learning , 2011, ArXiv.

[18]  Eyke Hüllermeier,et al.  Preferences in AI: An overview , 2011, Artif. Intell..

[19]  Tapio Salakoski,et al.  Learning intransitive reciprocal relations with kernel methods , 2010, Eur. J. Oper. Res..

[20]  Lurdes Y. T. Inoue,et al.  Decision Theory: Principles and Approaches , 2009 .

[21]  Michalis K. Titsias,et al.  Variational Learning of Inducing Variables in Sparse Gaussian Processes , 2009, AISTATS.

[22]  Manfred Opper,et al.  The Variational Gaussian Approximation Revisited , 2009, Neural Computation.

[23]  Kfir Eliaz,et al.  Indifference or indecisiveness? Choice-theoretic foundations of incomplete preferences , 2006, Games Econ. Behav..

[24]  Zoubin Ghahramani,et al.  Sparse Gaussian Processes using Pseudo-inputs , 2005, NIPS.

[25]  Carl E. Rasmussen,et al.  A Unifying View of Sparse Approximate Gaussian Process Regression , 2005, J. Mach. Learn. Res..

[26]  Wei Chu,et al.  Preference learning with Gaussian processes , 2005, ICML.

[27]  Alessandro Sperduti,et al.  Learning Preferences for Multiclass Problems , 2004, NIPS.

[28]  Dan Roth,et al.  Constraint Classification: A New Approach to Multiclass Classification , 2002, ALT.

[29]  Ren-Jye Yang,et al.  Metamodeling development for vehicle frontal impact simulation , 2001, DAC 2001.

[30]  Seth Rogers,et al.  Learning Subjective Functions with Large Margins , 2000, ICML.

[31]  Lothar Thiele,et al.  Comparison of Multiobjective Evolutionary Algorithms: Empirical Results , 2000, Evolutionary Computation.

[32]  David Barber,et al.  Bayesian Classification With Gaussian Processes , 1998, IEEE Trans. Pattern Anal. Mach. Intell..

[33]  Yoram Singer,et al.  Learning to Order Things , 1997, NIPS.

[34]  Hong Chang,et al.  Model Determination Using Predictive Distributions with Implementation via Sampling-Based Methods , 1992 .

[35]  Hervé Moulin,et al.  Choice functions over a finite set: A summary , 1985 .

[36]  Wayne Shafer,et al.  The Nontransitive Consumer , 1974 .

[37]  R. Luce Semiorders and a Theory of Utility Discrimination , 1956 .

[38]  S. Pfeifer A Course In Microeconomic Theory , 2016 .

[39]  Joseph B. Kadane,et al.  Coherent choice functions under uncertainty , 2009, Synthese.

[40]  Eyke Hüllermeier,et al.  Preference Learning , 2005, Künstliche Intell..

[41]  Marco Laumanns,et al.  Scalable Test Problems for Evolutionary Multiobjective Optimization , 2005, Evolutionary Multiobjective Optimization.

[42]  K. Obermayer,et al.  Learning Preference Relations for Information Retrieval , 1998 .

[43]  David J. C. MacKay,et al.  Bayesian Methods for Backpropagation Networks , 1996 .

[44]  L. Thurstone A law of comparative judgment. , 1994 .

[45]  Peter C. Fishburn,et al.  Nonlinear preference and utility theory , 1988 .

[46]  G. Debreu Mathematical Economics: Representation of a preference ordering by a numerical function , 1983 .