论文信息 - Data reduction: feature aggregation

Data reduction: feature aggregation

Feature aggregation is a process through which a set of new features is created, its purpose is improving performance such as estimated accuracy, visualization, and comprehensibility of learned knowledge. Feature aggregation is briefly reviewed in the framework of constructive induction and functional mapping. In the former we introduce basic operators for constructing new features and a typical algorithm; in the latter, we introduce some statistical methods and a neural network method.

Hiroshi Motoda | Huan Liu | H. Motoda | Huan Liu

[1] R. Michalski,et al. Multistrategy Constructive Induction: AQ17-MCI , 1993 .

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

[3] Hiroshi Motoda,et al. Feature Selection for Knowledge Discovery and Data Mining , 1998, The Springer International Series in Engineering and Computer Science.

[4] Peter Clark,et al. The CN2 Induction Algorithm , 1989, Machine Learning.

[5] Larry A. Rendell,et al. Constructive Induction On Decision Trees , 1989, IJCAI.

[6] Larry A. Rendell,et al. Constructive Induction Using Fragmentary Knowledge , 1996, ICML.

[7] H. Akaike. A new look at the statistical model identification , 1974 .

[8] J. Ross Quinlan,et al. Induction of Decision Trees , 1986, Machine Learning.

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

[10] P. Utgoff,et al. Multivariate Decision Trees , 1995, Machine Learning.

[11] Ryszard S. Michalski,et al. Hypothesis-Driven Constructive Induction in AQ17-HCI: A Method and Experiments , 1994, Machine Learning.

[12] Anthony J. Pettofrezzo,et al. Matrices and Transformations , 1978 .

[13] Gerhard Widmer,et al. Recognition and Exploitation of Contextual CLues via Incremental Meta-Learning , 1996, ICML.

[14] Christopher J. Matheus,et al. The Need for Constructive Induction , 1991, ML.