Knowledge Reduction Based on Oriented Relevance Information

YANG Li-xia,WEI Li-li
DOI: https://doi.org/10.3321/j.issn:1002-8331.2007.01.056
2007-01-01
Computer Engineering and Applications Journal
Abstract:Relevance information is a statistic which measures the non-linear relationships between two qualitative variables.In this paper,we define the oriented relevance information among attributes of information and decision systems,which measures the relevance among attributes of information and decision systems.And based on these definitions,we present a new discriminant theorem of knowledge reduction of information systems.Furthermore,our tests of information systems and inconsistent decision systems prove that the approach is feasible and efficient.
What problem does this paper attempt to address?