论文信息 - Techniques for Dealing with Missing Values in Classification

Techniques for Dealing with Missing Values in Classification

A brief overview of the history of the development of decision tree induction algorithms is followed by a review of techniques for dealing with missing attribute values in the operation of these methods. The technique of dynamic path generation is described in the context of tree-based classification methods. The waste of data which can result from casewise deletion of missing values in statistical algorithms is discussed and alternatives proposed.

[1] J. R. Quinlan. Discovering rules by induction from large collections of examples Intro-ductory readings in expert s , 1979 .

[2] Leo Breiman,et al. Classification and Regression Trees , 1984 .

[3] Earl B. Hunt,et al. Concept learning,: An information processing problem , 1974 .

[4] Philip J. Stone,et al. Experiments in induction , 1966 .

[5] Luc De Raedt,et al. Machine Learning: ECML-94 , 1994, Lecture Notes in Computer Science.

[6] A. P. White,et al. Probabilistic induction by dynamic part generation in virtual trees , 1987 .

[7] Trevor Hastie,et al. Statistical Models in S , 1991 .

[8] G. V. Kass. An Exploratory Technique for Investigating Large Quantities of Categorical Data , 1980 .

[9] Allan P. White,et al. Early Screening for Castric Cancer Using machine Learning Techniques , 1994, ECML.

[10] Allan P. White,et al. Machine learning techniques in early screening for gastric and oesophageal cancer , 1996, Artif. Intell. Medicine.

[11] Donald Michie,et al. Expert systems in the micro-electronic age , 1979 .

[12] Allan P. White,et al. The importance of attribute selection measures in decision tree induction , 2005, Machine Learning.

[13] Daryl Pregibon,et al. Tree-based models , 1992 .

[14] Ivan Bratko,et al. Experiments in automatic learning of medical diagnostic rules , 1984 .

[15] Ron Kohavi,et al. Lazy Decision Trees , 1996, AAAI/IAAI, Vol. 1.