DOI: 10.14489/vkit.2014.07.pp.056-063

Пушкарев А. Н.
МАТЕМАТИЧЕСКОЕ ОБЕСПЕЧЕНИЕ МАРКЕТИНГОВОЙ СИСТЕМЫ ПОДДЕРЖКИ ПРИНЯТИЯ РЕШЕНИЙ
(c. 56-63)

Аннотация. Рассмотрены подходы к математическому моделированию процесса выбора объекта экспертом. Приведено описание моделей, разработанных в рамках предложенных подходов. Примеры прототипа маркетинговой системы поддержки принятия решений иллюстрируют практическую значимость работы. Приложение обеспечивает проведение имитационного эксперимента с использованием созданного математического аппарата.

Ключевые слова: маркетинг; объект; эксперт; характеристика; процесс выбора; математическое моделирование; система поддержки принятия решений.

Pushkarev A. N.
MATHEMATICAL SUPPORT FOR MARKETING DECISION SUPPORT SYSTEM
(pp. 56-63)

Abstract. The article considers approaches to the mathematical modeling of object selection by an expert. They are based on a more general approach, which represents the object selection as a sequence of successive states. Each state represents a comparison of objects for some of their characteristics. The process ends with a transition to a state in which the expert prefers one of the objects or rejects them all. In the first approach, developed by the author, the modeling of the selection process is carried out by means of the mathematical appa-ratus of the theory of Markov chains. The author gives a description of two models of object selection by an expert, developed within the frame-work of the proposed approach. Cursory selection model assumes that the expert examines the characteristics of objects in an arbitrary order and has the opportunity to return to the characteristic he examined earlier. In this model the expert makes a decision based on the information about the latest examined characteristic of objects. Methodical selection model suggests that the expert examines the characteristics of objects in descending order of their importance and without return to the characteristic, he examined earlier. This model takes into account the impact of all examined characteristics on the expert’s decision. To develop a model that takes into account the contributions of all characteristics examined by an expert in an arbitrary order, the author proposed an approach to the selection process modeling based on probability theory and combinatorial analysis. Methodless selection model created within the framework of this approach considers the total probabilistic contribution of all possible sequences of characteristics in the expert’s decision. The practical significance of the work is illustrated by the prototype of a marketing decision support system. The application provides the carrying out of simulation experiment using the created mathematical apparatus. The input parameters of the system are the values of importance and expert satisfaction with the characteristics of objects. Output data are point and interval estimates of the probabilities that expert will select one of the objects or reject them all.

Keywords: Marketing; Object; Expert; Characteristic; Selection process; Mathematical modeling; Decision support system.