Supervised learning rule selection for multiclass decision with performance constraints

Nisrine Jrad (1) , Edith Grall-Maës (1) , Pierre Beauseroy (1)

1 LM2S - Laboratoire Modélisation et Sûreté des Systèmes

A procedure to select a supervised rule for multiclass problem from a labeled dataset is proposed. The rule allows class-selective rejection and performance constraints. The unknown probabilities are estimated with a Parzen estimator. A set of rules are built by varying the Parzen¿s smoothness parameter of the marginal probabilities estimates and plugging them into the statistical hypothesis rules. A criterion that assesses the quality of these rules is estimated and used to select a rule. Resampling and aggregation methods are used to show the efficiency of the estimated criterion.

Domains: Computer Science [cs]; Computer Vision and Pattern Recognition [cs.CV]

Nisrine Jrad, Edith Grall-Maës, Pierre Beauseroy. Supervised learning rule selection for multiclass decision with performance constraints. ICPR 2008 19th International Conference on Pattern Recognition, Dec 2008, Tampa, United States. pp.1-4, ⟨10.1109/ICPR.2008.4761200⟩. ⟨hal-02362961⟩