The decision-making algorithm selection is one of the problems that modern economics, industrial management and situation identification process face. The Bayesian composite approach is considered to be the safest way for algorithm identification. However, the approach based on Bayesian composite rules is time-consuming and computationally intensive, that`s why it is important to develop the method, when combined with the Bayesian approach, will enable to scale back the computation. The article describes the synthesis method of flow graphs for optimal search within the member group (including the search within the industrial structure) and suggests this method as the main implementable algorithm for situation identification, based on Bayesian composite rule for decision-making. Such scientific methods as mathematical techniques of graph theory, algorithm and combination methods, probability theory method, and methods of statistical analysis were used in the research.

The article describes the development of multifaceted and efficient approaches to the context information analysis for synthesis of industrial situations context recognition algorithm in automated management systems within the enterprises. The probability theory method and method of statistical analysis, decision theory method, methods of algorithm and combination theory were used while researching. The research resulted in the development of new approaches to the context information analysis framework for pattern recognition which enables us to identify the procedure of contextual recognition for synthesis of working industrial situation recognition algorithm. A correspondence between the recognition error rate and the guaranteed recognition threshold, which can be used for setting up the automated context-based recognition systems, was analytically obtained during the research.