Learning acyclic decision trees with Functional Dependency Network and MDL Genetic Programming   [DT] [GP]

by

Shum, W.-H., Leung, K.-S. and Wong, M.-L.

Literature search on Evolutionary ComputationBBase ©1999-2013, Rasmus K. Ursem
     Home · Search · Adv. search · Authors · Login · Add entries   Webmaster
Note to authors: Please submit your bibliography and contact information - online papers are more frequently cited.

Info: International Multi-Conference on Computing in the Global Information Technology, ICCGI '06 (Conference proceedings), 2006, p. 25
Keywords:genetic algorithms, genetic programming
Abstract:
One objective of data mining [DM] is to discover parent-child relationships among a set of variables in the domain. Moreover, showing parents' importance can further help to improve decision makings' quality. Bayesian network [BN] (BN) is a useful model for multi-class problems and can illustrate parent-child relationships with no cycle. But it cannot show parents' importance. In contrast, decision trees state [DT] parents' importance clearly, for instance, the most important parent is put in the first level. However, decision trees [DT] are proposed for single-class problems only, when they are applied to multi-class ones, they are likely to produce cycles representing tautologic. In this paper, we propose to use MDL genetic programming [GP] (MDLGP) and functional dependency network (FDN) to learn a set of acyclic decision trees [DT] (Shum et al., 2005). The FDN is an extension of BN; it can handle all of discrete, continuous, interval and ordinal values; it guarantees to produce decision trees [DT] with no cycle; its learning search space [SS] is smaller than decision trees'; and it can represent higher-order relationships among variables. The MDLGP is a robust genetic programming (GP) [GP] [GPG] proposed to learn the FDN. We also propose a method to derive acyclic decision trees from [DT] the FDN. The experimental results demonstrate that the proposed method can successfully discover the target decision trees, [DT] which have no cycle and have the accurate classification results
Notes:
Dept. of Comput. Sci. & Eng., Chinese Univ. of Hong Kong
Internet search:Search Google
Search Google Scholar
Search Citeseer using Google
Search Google for PDF
Search Google Scholar for PDF
Search Citeseer for PDF using Google

Review item:

Mark as doublet (will be reviewed)

Print entry




BibTex:
@InProceedings{Shum:2006:ICCGI,
  author =       "Wing-Ho Shum and Kwong-Sak Leung and Man-Leung Wong",
  title =        "Learning acyclic decision trees with Functional
                 Dependency Network and {MDL} Genetic Programming",
  booktitle =    "International Multi-Conference on Computing in the
                 Global Information Technology, ICCGI '06",
  year =         "2006",
  pages =        "25",
  address =      "Bucharest",
  month =        "1-3 " # aug,
  publisher =    "IEEE",
  keywords =     "genetic algorithms, genetic programming",
  ISBN =         "0-7695-2690-X",
  doi =          "doi:10.1109/ICCGI.2006.46",
  abstract =     "One objective of data mining is to discover
                 parent-child relationships among a set of variables in
                 the domain. Moreover, showing parents' importance can
                 further help to improve decision makings' quality.
                 Bayesian network (BN) is a useful model for multi-class
                 problems and can illustrate parent-child relationships
                 with no cycle. But it cannot show parents' importance.
                 In contrast, decision trees state parents' importance
                 clearly, for instance, the most important parent is put
                 in the first level. However, decision trees are
                 proposed for single-class problems only, when they are
                 applied to multi-class ones, they are likely to produce
                 cycles representing tautologic. In this paper, we
                 propose to use MDL genetic programming (MDLGP) and
                 functional dependency network (FDN) to learn a set of
                 acyclic decision trees (Shum et al., 2005). The FDN is
                 an extension of BN; it can handle all of discrete,
                 continuous, interval and ordinal values; it guarantees
                 to produce decision trees with no cycle; its learning
                 search space is smaller than decision trees'; and it
                 can represent higher-order relationships among
                 variables. The MDLGP is a robust genetic programming
                 (GP) proposed to learn the FDN. We also propose a
                 method to derive acyclic decision trees from the FDN.
                 The experimental results demonstrate that the proposed
                 method can successfully discover the target decision
                 trees, which have no cycle and have the accurate
                 classification results",
  notes =        "Dept. of Comput. Sci. & Eng., Chinese Univ. of Hong
                 Kong",
}