A New Approach to Computing Weighted Attributes Values in Incomplete Information Systems

The Rough Sets theory has been conceived to deal with inexact, uncertain or vague knowledge in Information systems. The article is devoted to propose notions including the generalized discernibility matrix, the missing attribute set of an object, the indiscernible object set of an object, and the missing object set of an information system. A new algorithm for computing weighted attributes values of incomplete information systems is presented. This algorithm can reveal weighted attributes values and rules contained from an original given data alone with small computational cost. Therefore it is suitable to be used in the data preprocessing stage of knowledge discovery process. An example is given in the paper to show that the proposed algorithm well foresees real value of missing attribute and outperforms the methods based on ignoring missing values.

• Type: Journal
• Year: 2007
• Where: JCIT
• Authors: Tian Hong, Wang Xiukun, Henryk Rybinski
• Keywords: Generalized Discernibility Matrix, Information Systems, JCIT 2007, Rough Sets Theory