Stochastically informative expertise as natural step in expert systems net development

Abstract

At this paper is analyzed the problem of expert systems net development and stochastically informative expertise systems, as a very important component of the net, preparing. Specifically, the work considers the case, when for the feature possibilities expertise is used the distribution possibilities of examined feature probability. The necessity of these methods is reasoned that calculating index, choosing of single-or ranged estimates from a variety of assumptions, later attempt to restore the environment, which would have allowed to see how a distribution possibilities of examined feature probability. The paper presents the concept of the evaluation and the practical scheme realization rating, while analyzing the situation where the expert way of measuring revenues from company newly developed product sales opportunities in the three markets, coupled with the company the best distribution of sales in selected markets.