Digitale Bibliotheek
Sluiten Bladeren door artikelen uit een tijdschrift
 
<< vorige    volgende >>
     Tijdschrift beschrijving
       Alle jaargangen van het bijbehorende tijdschrift
         Alle afleveringen van het bijbehorende jaargang
           Alle artikelen van de bijbehorende aflevering
                                       Details van artikel 4 van 6 gevonden artikelen
 
 
  IGLUE: A lattice-based constructive induction system
 
 
Titel: IGLUE: A lattice-based constructive induction system
Auteur: Engelbert Mephu Nguifo
Patrick Njiwoua
Verschenen in: Intelligent data analysis
Paginering: Jaargang 5 (2001) nr. 1 pagina's 73-91
Jaar: 2001-04-01
Inhoud: A machine learning (ML) system which combines lattice-based and instance-based learning (IBL) techniques, is motivated and developed in this paper. We describe an IBL system over lattice theory called IGLUE that significantly improved both the complexity and accuracy of lattice-based learning systems. For this purpose, IGLUE uses the entropy function to select relevant lattice nodes, then extracts a set of new numerical features from the original set of boolean features, and finally applies a nearest neighbor technique with the Mahalanobis distance as the similarity measure between redescribed instances. IGLUE treats only domains described with symbolic features. In this paper, we present results of experiments we carried out to assess how well IGLUE performs on real problems, with other similarity measures and selection functions. We combine three selection functions and three similarity measures, and thus run nine experiments. We compare the performance of these combined strategies on a collection of ML benchmarks. Empirical results indicate that IGLUE is able to achieve good classification accuracy in a variety of domains, whatever the selection function or the similarity measure mentioned above. These new functions and measures highlight the importance of instance-based learning through lattice theory.
Uitgever: IOS Press
Bronbestand: Elektronische Wetenschappelijke Tijdschriften
 
 

                             Details van artikel 4 van 6 gevonden artikelen
 
<< vorige    volgende >>
 
 Koninklijke Bibliotheek - Nationale Bibliotheek van Nederland