Skip to Main content Skip to Navigation
Conference papers

Variable nodale et cône implicatif

Régis Gras 1 Dominique Lahanier-Reuter 2 M. Bailleul
2 Théodile-CIREL
CIREL - Centre Interuniversitaire de Recherche en Education de Lille (CIREL) - EA 4354 : EA1764
Abstract : Inside the Statistical Implicative Analysis frame, implicative rules may be extracted while pondered by a measure called Intensity of Implication. The set of the non-symmetrical relationships that associate the different variables may be represented by a graph so called implicative. Inside the graph, relationships upstream and downstream from a central variable parted on either sides of this node can be sometimes noticed. This situation of a confluent is metaphorically said a two clusters implicative cone. On condition that the set is rather homogeneous and linked, giving conceptual meaning to this cone allows some kind of causalities/consequences isolation of it from the whole set. In this communication, we define criteria for the homogeneity of a cone and conditions for existence of causal relationships. Some significant examples illustrate our theoretical discourse.
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-01258157
Contributor : Lina Duke <>
Submitted on : Monday, January 18, 2016 - 4:55:15 PM
Last modification on : Tuesday, April 21, 2020 - 9:33:27 AM

Identifiers

  • HAL Id : hal-01258157, version 1

Collections

Citation

Régis Gras, Dominique Lahanier-Reuter, M. Bailleul. Variable nodale et cône implicatif. 8 ème colloque international sur l'Analyse Statistique Implicative, des sciences dures aux sciences humaines et sociales - ASI 8, Nov 2015, Radès, Tunisie. pp.81-98. ⟨hal-01258157⟩

Share

Metrics

Record views

175