A Distance-Based Attribute Selection Measure for Decision Tree Induction
关键词:
摘要: This note introduces a new attribute selection measure for ID3-like inductive algorithms. is based on distance between partitions such that the selected in node induces partition which closest to correct of subset training examples corresponding this node. The relationship with Quinlan's information gain also established. It formally proved our not biased towards attributes large numbers values. Experimental studies confirm previously reported results showing predictive accuracy induced decision trees sensitive goodness measure. However, produces smaller than ratio Quinlan, especially case data whose have significantly different
springer.com
acm.org
paperity.org参考文章(6)
Ivan Bratko, Igor Kononenko, Bojan Cestnik, ASSISTANT 86: a knowledge-elicitation tool for sophisticated users EWSL'87 Proceedings of the 2nd European Conference on European Working Session on Learning. pp. 31- 45 ,(1987)
Tim Niblett, Peter Clark, Induction in noisy domains EWSL'87 Proceedings of the 2nd European Conference on European Working Session on Learning. pp. 11- 30 ,(1987)
John Mingers, An Empirical Comparison of Selection Measures for Decision-Tree Induction Machine Learning. ,vol. 3, pp. 319- 342 ,(1989) , 10.1023/A:1022645801436
J.R. Quinlan, Induction of Decision Trees Machine Learning. ,vol. 1, pp. 81- 106 ,(1986) , 10.1023/A:1022643204877
Ramon Lopez de Mantaras Badia, Autoapprentissage d'une partition : application au classement itératif de données multidimensionnelles Université Paul Sabatier - Toulouse III. ,(1977)
Ivan Bratko, Igor Kononenko, E. Roskar, Experiments in automatic learning of medical diagnostic rules ,(1984)