Data Exploration Based on Local Attribution Explanation: A Medical Use Case

Exploration de données basée sur les Explications locales attributives : un exemple d'utilisation médical

Elodie Escriva (1, 2, 3) , Emmanuel Doumard (2, 4) , Jean-Baptiste Excoffier (1) , Julien Aligon (2, 3) , Paul Monsarrat (4, 5, 6) , Chantal Soulé-Dupuy (2, 3)

1 Kaduceo
2 IRIT-SIG - Systèmes d’Informations Généralisées
3 UT Capitole - Université Toulouse Capitole
4 RESTORE - Geroscience and rejuvenation research center
5 Service Oto-Rhino-Laryngologie (ORL) et chirurgie cervico-faciale [CHU Toulouse]
6 ANITI - Artificial and Natural Intelligence Toulouse Institute

Exploratory data analysis allows to discover knowledge and patterns and to test hypotheses. Modelling predictive tools associated with explainability made it possible to explore more and more complex relationships between attributes. This study presents a method to use local explanations as a new data space to retrieve precise and pertinent information. We aim to apply this method to a medical dataset and underline the benefit of using explanations to gain knowledge. In particular, we show that clusters based on local explanations, combined with decision rules, allow to better characterise patient subgroups.

Mots clés: Medical Data Exploration; Explainable IA; Machine Learing

Domaines: Apprentissage [cs.LG]; Analyse de données, Statistiques et Probabilités [physics.data-an]; Médecine humaine et pathologie

Elodie Escriva, Emmanuel Doumard, Jean-Baptiste Excoffier, Julien Aligon, Paul Monsarrat, et al.. Data Exploration Based on Local Attribution Explanation: A Medical Use Case. 27th European Conference Advances in Databases and Information Systems (ADBIS 2023), Sep 2023, Barcelone, Spain. pp.315-323, ⟨10.1007/978-3-031-42941-5_27⟩. ⟨hal-04239173⟩