Identification of Integrated Rating Mechanisms as Optimization Problem

Abstract

A problem of identification of an integrated ranking mechanism is considered. We present a perspective approach to the task of identification of an integrated ranking mechanism - we show that representation of any such mechanism in one-hot encoded form allows us to formulate such a task as an optimization problem that may be solved with a wide variety of optimization techniques. The approach suggested is illustrated by several examples of identification or approximation of learning sets generated by several Boolean functions.