Synthesis of the optimal mechanism of active expertise: realization via computer modeling systems

The paper considers a model of active expertise as expertise that takes into account the experts' possible interest in its results. We study its one-dimensional formulation — the problem of one-criterion collective choice or, in other terminology, active planning. There is a goal procedure for expertise that is optimal when experts are not active. The problem of synthesis of the optimal mechanism of active expertise is given. The article presents proof of solution existence to the problem of synthesis of the optimal mechanism of one-dimensional active expertise. Earlier, only an algorithm for its obtaining by solving a system of equations was proposed. We show that, within the framework of generally accepted assumptions on the class of goal procedures, a solution to the problem of active planning always exists and is so-called generalized median voter schemes, described by Herve Moulin in 1980 as a class of strategy-proof collective choice rules (or mechanisms in the terminology of the theory of active systems), which are Pareto efficient. A model for analyzing mechanisms of active expertise based on a GMVS and a model for synthesizing an optimal strategy-proof mechanism have been developed in the simulation environment. A search for the parameters of the equivalent direct mechanism is also implemented.