Représentativité, généricité et singularité : augmentation de données pour l’exploration de dossiers médicaux

Abstract

In order to classify individuals according to exemplars that represent them accurately, data visualisation applied to the medical field need to avoid overgeneralization : each case must be treated as a particular instance. This paper presents an algorithm allowing each individual in a dataset to rank other individuals and vote for those that match their important features. Aggregating all these votes give us a way to visualize data according to typical individuals representing subsets of closely-related patients.