A generaliser applied to SBL

In this work two algorithms have been developed: a generaliser and a hierarchy constructor. The generaliser receives as inputs the descriptions of some positive examples of a concept and it calculates their maximal conjunctive generalisation. This algorithm is an important part within the learning a...

Descripción completa

Detalles Bibliográficos
Autor: Moreno Ribas, Antonio
Tipo de recurso: informe técnico
Fecha de publicación:1992
País:España
Institución:Universitat Politècnica de Catalunya (UPC)
Repositorio:UPCommons. Portal del coneixement obert de la UPC
Idioma:inglés
OAI Identifier:oai:upcommons.upc.edu:2117/370354
Acceso en línea:https://hdl.handle.net/2117/370354
Access Level:acceso abierto
Palabra clave:Graph algorithms
Computer algorithms
Algorismes de grafs
Algorismes computacionals
Àrees temàtiques de la UPC::Informàtica
Descripción
Sumario:In this work two algorithms have been developed: a generaliser and a hierarchy constructor. The generaliser receives as inputs the descriptions of some positive examples of a concept and it calculates their maximal conjunctive generalisation. This algorithm is an important part within the learning algorithm proposed by Gustavo Núñez, which is based on the SBL (Similitude Based Learning) paradigm. The hierarchy constructor receives the generalisations calculated by the first algorithm and integrates them within a graph structure, which expresses the relationships between the concepts described by the generalisations. Both algorithms are written in Common Lisp and have been tested using the blocks world domain, but the generaliser has been written in such a way that it can be used in any domain satisfying some restrictions.