Decision tree induction methods using an entropy criterion – I. Global approaches - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 1997

Decision tree induction methods using an entropy criterion – I. Global approaches

Résumé

A great number of systems can only be described by models established through the statistical analysis of empirical data. When both the output and the input data are qualitative, automatic learning algorithms have to be designed. Those algorithms generate rules which describe the input/output behaviour of the system, from the observed data. In this contribution, we propose to generate non deterministic rules from a set of incoherent data. We first introduce an information based index in order to measure the incoherence of the learning set (the feasibility of the modelling problem is thus measured). Then, we propose a procedure which allows to increase the coherence of the learning set, by considering only a subset for which the feasibility index is greater. The determination of such a subset is made in an optimal way, such that the lost information is minimized. In order to construct a simpler model, we search a set of explanatory variables such as its explanation power is great : the explanation power of each variable taken separately has to be important, and the common information brought by all the parts of this set has to be low. We present four « by level » approaches based on the construction of a decision tree. Among these approaches, we distinguish the bottom-up approach and the top-down approach, associated to a merging or splitting algorithm. In order to compare them, we apply these methods on several databases.
Fichier non déposé

Dates et versions

hal-01509913 , version 1 (18-04-2017)

Identifiants

  • HAL Id : hal-01509913 , version 1

Citer

Paul-Benoît Perche, Denis Pomorski. Decision tree induction methods using an entropy criterion – I. Global approaches. Second International ICSC (International Computer Science Conventions), Symposium on “Soft Computing” (SOCO’97), Sep 1997, Nimes, France. pp.286-293. ⟨hal-01509913⟩

Collections

CNRS LAGIS
31 Consultations
0 Téléchargements

Partager

Gmail Facebook X LinkedIn More