Robust supervised classification with mixture models: Learning from data with uncertain labels

Charles Bouveyron (1, 2) , Stéphane Girard (3)

1 CES - Centre d'économie de la Sorbonne
2 SAMOS - Statistique Appliquée et MOdélisation Stochastique
3 MISTIS - Modelling and Inference of Complex and Structured Stochastic Systems

In the supervised classification framework, human supervision is required for labeling a set of learning data which are then used for building the classifier. However, in many applications, human supervision is either imprecise, difficult or expensive. In this paper, the problem of learning a supervised multi-class classifier from data with uncertain labels is considered and a model-based classification method is proposed to solve it. The idea of the proposed method is to confront an unsupervised modelling of the data with the supervised information carried by the labels of the learning data in order to detect inconsistencies. The method is able afterward to build a robust classifier taking into account the detected inconsistencies into the labels. Experiments on artificial and real data are provided to highlight the main features of the proposed method as well as an application to object recognition under weak supervision.

Keywords: robustness; label noise; weakly-supervised classification; supervised classification; data with uncertain labels; mixture models

Domains: Statistics [math.ST]; Statistics Theory [stat.TH]

Charles Bouveyron, Stéphane Girard. Robust supervised classification with mixture models: Learning from data with uncertain labels. Pattern Recognition, 2009, 42 (11), pp.2649-2658. ⟨10.1016/j.patcog.2009.03.027⟩. ⟨hal-00325263v2⟩