On a Possibility of Applying Interrelationship Mining to Gene Expression Data Analysis

Decision logic for rough set-based interrelationship mining

“…As we have discussed in [3], [4], we assumed that binary relations between values of different attributes are given. From a viewpoint of applying interrelationship mining to data analysis, it is very important to (semi-)automatically detect or decide the binary relations for seeking interrelationships between attributes.…”

“…This reformulation is based on revising the set R AT of families of binary relations for comparing attribute values and expression of interrelationships by new condition attributes [4].…”

“…For each attribute a ∈ AT , each value v ∈ V a , and each binary relation R a in S, a triple (a, R a , v) is called an atomic formula or a descriptor over AT , R AT , and V for low order decision rules. We then introduce the set of atomic formulas A low for low order decision rules as follows: (4) where {R a } is the family of binary relations defined on V a .…”

“…In this section, we review a concept of interrelationship mining [2], [3], [4]. One of the main motivations of interrelationship mining is to extract characteristics based on interrelationship between different attributes such that:…”

“…The authors have proposed a concept of interrelationship mining [2], [3], [4] that enables us to extract characteristics based on comparison between values of different attributes. In this paper, we consider logical aspects of interrelationship mining by introducing decision logic, in particular, decision languages and those semantics, and low and high order decision rules for interrelationship mining.…”