Solving the Choice Problem in the Conditions of Physical Diversity and Limited Observable Signs
Abstract
Purpose: The diversity of the input information and lack of methods to amalgamate miscellaneous information for the choice problem solution call for the development of an appropriate technique. The purpose of this work is developing methods to solve the choice problem taking into account the diversity and limited nature of the received information. Results: A mathematical definition has been formulated for the choice problem. The task is reduced to checking the multidimensional statistical hypothesis that selective signs vectors are equal to the reference values. To solve this task, the credibility relation criterion is used which is the most powerful one. A combined proximity indicator is introduced which is a measure of proximity to the reference values for all the selective signs of the object received by measurements. A technique for the choice problem solution is discussed for a sufficiently large volume of the information on the objects. In actual practice, the volume of the measuring information on the characteristics of the observed objects is usually limited. In this case, Pearson's distribution of the 3rd type approximates the true distribution of the combined proximity indicator well enough. A technique to solve the choice problem with a limited amount of information is developed. Practical relevance: Using the developed methods will help to increase the reliability of the choice problem as compared to the existing methods.Published
2015-08-01
How to Cite
Arseniev, V., & Trofimov, I. (2015). Solving the Choice Problem in the Conditions of Physical Diversity and Limited Observable Signs. Information and Control Systems, (4), 114-118. Retrieved from http://proceedings.spiiras.nw.ru/index.php/ius/article/view/4362
Issue
Section
Information and measuring systems