Per citar aquest document: http://ddd.uab.cat/record/153511
A defeasible reasoning model of inductive concept learning from examples and communication
Ontañón Villar, Santi (Institut d'Investigació en Intel·ligència Artificial)
Dellunde i Clavé, Pilar (Universitat Autònoma de Barcelona)
Godo, Luís (Institut d'Investigació en Intel·ligència Artificial)
Plaza, Enric (Institut d'Investigació en Intel·ligència Artificial)

Data: 2012
Resum: This paper introduces a logical model of inductive generalization, and specifically of the machine learning task of inductive concept learning (ICL). We argue that some inductive processes, like ICL, can be seen as a form of defeasible reasoning. We define a consequence relation characterizing which hypotheses can be induced from given sets of examples, and study its properties, showing they correspond to a rather well-behaved non-monotonic logic. We will also show that with the addition of a preference relation on inductive theories we can characterize the inductive bias of ICL algorithms. The second part of the paper shows how this logical characterization of inductive generalization can be integrated with another form of non-monotonic reasoning (argumentation), to define a model of multiagent ICL. This integration allows two or more agents to learn, in a consistent way, both from induction and from arguments used in the communication between them. We show that the inductive theories achieved by multiagent induction plus argumentation are sound, i. e. they are precisely the same as the inductive theories built by a single agent with all data.
Drets: Tots els drets reservats.
Llengua: Anglès
Document: article ; recerca ; submitedVersion
Matèria: Induction ; Logic ; Argumentation ; Machine learning ; Concept learning
Publicat a: Artificial intelligence, Vol. 193 (2012) , p. 129-148, ISSN 0004-3702

DOI: doi:10.1016/j.artint.2012.08.006


Pre-print
43 p, 761.1 KB

El registre apareix a les col·leccions:
Articles > Articles de recerca
Articles > Articles publicats

 Registre creat el 2016-05-19, darrera modificació el 2016-06-12



   Favorit i Compartir