Fuzzy decision trees and numerical attributes
暂无分享,去创建一个
Classical crisp decision trees (DT) are widely applied to classification tasks. Nevertheless, there are also a lot of fuzzy decision tree (FDT) solutions. Sometimes fuzzy borders for discretisation of continuous-valued attributes are used. The induction of FDT in these solutions need some parameters from the expert or user. An automatically generated membership function for discretisation of continuous-valued attributes is described in this paper. An example with the decision tree construction and the unseen data classification is given. The results of crisp and fuzzy decision trees are compared at the end.
[1] Günter Seidelmann. Using Heuristics to Speed up Induction on Continuous-Valued Attributes , 1993, ECML.
[2] Xindong Wu,et al. Fuzzy Interpretation of Induction Results , 1995, KDD.
[3] I. Hatono,et al. Fuzzy decision trees by fuzzy ID3 algorithm and its application to diagnosis systems , 1994, Proceedings of 1994 IEEE 3rd International Fuzzy Systems Conference.
[4] Usama M. Fayyad,et al. Multi-Interval Discretization of Continuous-Valued Attributes for Classification Learning , 1993, IJCAI.