Query-based learning of acyclic conditional preference networks from noisy data

Fabien Labernia (1) , Florian Yger (1) , Brice Mayag (1) , Jamal Atif (1)

1 LAMSADE - Laboratoire d'analyse et modélisation de systèmes pour l'aide à la décision

Conditional preference networks (CP-nets) provide a powerful, compact, and intuitive graphical tool to represent the preferences of a user. However learning such a structure is known to be a difficult problem due to its combinatorial nature. We propose in this paper a new, efficient, and robust query-based learning algorithm for acyclic CP-nets. In particular, our algorithm takes into account the incoherences in the user’s preferences or in noisy data by searching in a principled way the variables that condition the other ones. We provide complexity results of the algorithm, and demonstrate its efficiency through an empirical evaluation on synthetic and on real datasets.

Mots clés: Preference learning; machine learning; artificial intelligence

Domaines: Intelligence artificielle [cs.AI]

Fabien Labernia, Florian Yger, Brice Mayag, Jamal Atif. Query-based learning of acyclic conditional preference networks from noisy data. EURO Mini Conference: "From Multiple Criteria Decision Aid to Preference Learning" (DA2PL'2016), Nov 2016, Paderborn, Germany. pp.6. ⟨hal-01461579⟩