Uncertain variables as a tool for analysis and decision making in uncertain knowledge-based systems

摘要

Purpose - To present new results concerning the applications of the uncertain variables to analysis and decision problems in a class of uncertain knowledge-based systems described by a relational knowledge representation. Design/methodology/approach - The analysis and decision-making problems are formulated, the general forms of the solutions are described, the theorem concerning the optimal solution is presented, numerical examples illustrating the general theoretical approaches are included. Findings - The uncertain variables have been proved as convenient tool for solving specific optimization problems based on relational knowledge representations with unknown parameters. Originality/value - Formulation and solving optimization problem for the given certainty threshold, application to the problem of resource and task distributions in a complex of parallel operations, indications of possible practical applications.