Multigranulation Supertrust Model for Attribute Reduction

As big data often contains a significant amount of uncertain, unstructured, and imprecise that are structurally complex incomplete, traditional attribute reduction methods less effective when applied to large-scale incomplete information systems extract knowledge. Multigranular computing provides powerful tool for use in analysis conducted at different levels granularity. In this article, we present novel multigranulation supertrust fuzzy-rough set-based (MSFAR) algorithm support the formation hierarchies granules higher types orders, which addresses newly emerging mining problems analysis. First, model based on valued tolerance relation is constructed identify fuzzy similarity changing knowledge granularity with multimodality attributes. Second, an ensemble consensus compensatory scheme was adopted calculate multigranular trust degree reputation granularities create reasonable subproblems granulation levels. Third, equilibrium method coevolution employed ensure wide range balancing exploration exploitation, strategy can classify super elitists' preferences detect noncooperative behaviors global convergence ability high search accuracy. The experimental results demonstrate MSFAR achieves performance addressing uncertain large number multigranularity variables.