Y. Bao, Y. Lu, L. Sun, and J. Zhang (PRC)
Discretization; Feature selection; Rough sets; Evaluation
Discretization and feature selection are effective techniques in handling continuous attributes and improving efficiency for data mining. Most Discretization methods are local and easy to lose valuable information under the situation of small samples. Moreover, features gathered to identify objects are often not all useful or equally informative and some of them may be redundant or irrelevant. In this paper, a hybrid and rough sets based method with inconsistency checking is presented to facilitate the discretization for continuous attributes and feature selection in evaluation. After a quick review on basic notions of rough sets, general procedures of the method are illustrated by an example, which shows the method's validity.
Important Links:
Go Back
