Cautious label ranking with label-wise decomposition

Sébastien Destercke (1, 2) , Marie-Hélène Masson (1, 2) , Michael Poss (3)

1 Labex MS2T - Laboratoire d'Excellence "Maîtrise des Systèmes de Systèmes Technologiques"
2 Heudiasyc - Heuristique et Diagnostic des Systèmes Complexes [Compiègne]
3 MAORE - Methods, Algorithms for Operations REsearch

In this paper, we are interested in the label ranking problem. We are more specifically interested in the recent trend consisting in predicting partial but more accurate (i. e. , making less incorrect statements) orders rather than complete ones. To do so, we propose a ranking method based on label-wise decomposition. We estimate an imprecise probabilistic model over each label rank and we infer a partial order from it using optimisation techniques. This leads to new results concerning a particular bilinear assignment problem. Finally, we provide some experiments showing the feasibility of our method.

Mots clés: Bi; assignment problem; label - wise decomposition; Label ranking; imprecise probability; partial predictions; linear optimisation

Domaines: Intelligence artificielle [cs.AI]; Apprentissage [cs.LG]

Sébastien Destercke, Marie-Hélène Masson, Michael Poss. Cautious label ranking with label-wise decomposition. European Journal of Operational Research, 2015, 246 (3), pp.927-935. ⟨10.1016/j.ejor.2015.05.005⟩. ⟨hal-01166139⟩