Разработка подсистемы работы с байесовскими сетями для системы поддержки принятия решений в управлении проектами по разработке программного обеспечения

Abstract

Bayesian networks are currently a popular tool for solving various problems, including creating decision support systems. This paper proposes a tool for creating a Bayesian network and direct probabilistic inference. The specificity of the problem consists in working with large networks (more than 1000 nodes) with a large number of parent nodes at one node (15 or more). The tool is integrated with Redmine and allows you to calculate the probability of a manager's error when determining the current state of the project.