[GNT14] A simple and robust scoring technique for binary classification

Revue Internationale avec comité de lecture : Journal Artificial Intelligence Research, vol. 3(1), pp. 52-58, 2014, (doi:10.5430/air.v3n1p52)

Mots clés: Binary outcome, Prediction, Scores, Small learning set

Résumé: A new simple scoring technique is developed in a binary supervised classification context when only a few observations are available. It consists in two steps: in the first one partial scores are obtained, one for each predictor, either categorical or continuous. Each partial score is a discrete variable with 7 values ranging from -3 to 3, based upon an empirical comparison of the distributions for each class. In a second step the partial scores are added and standardised into a global score, which allows a decision rule. This simple technique is successfully compared with classical supervised techniques for a classical benchmark and has been proved to be especially well fitted in an industrial problem.

Equipe: msdma


@article {
title="{A simple and robust scoring technique for binary classification}",
author="C. Gomes and H. Nocairi and M. Thomas and J. Collin and G. Saporta",
journal="Artificial Intelligence Research",