Attributive and Object Subcontexts in Inferring Good Maximally Redundant Tests
Xenia Aleksandrovna Naidenova, Vladimir Andreevich Parkhomenko · 2014
Abstract. Inferring Good Maximally Redundant Classification Tests (GMRTs) as Formal Concepts is considered. Two kinds of classification subcontexts are defined: attributive and object ones. The rules of forming and reducing subcontexts based on the notion of essential attributes and objects are given. They lead to the possibility of the inferring control. In particular, an improved Algorithm for Searching all GMRTs on the basis of attributive subtask is proposed. The hybrid attributive and object approaches are presented. Some computational aspects of algorithms are analyzed.