The article addresses the synthesis of the computing core of an intelligent internal audit support system (IIASS) for a business entity. A formalized mathematical model of the IIASS is developed, focusing on the assessment of economic risks that may affect a company's financial stability, profitability, and operational efficiency. Unlike most existing approaches that rely on manual audit procedures or fragmented statistical methods, the proposed unified model integrates logical, probabilistic, and fuzzy components within a single intelligent environment. The audit object feature space is examined as a foundation for the application of machine learning methods and rule-based logic in automated risk evaluation. A risk function is constructed to model the criticality level of internal audit objects through a parameterized combination of features, incorporating feature weighting and an adaptive decision threshold mechanism. A hybrid architecture for the IIASS is proposed, combining rule-based logic, logistic regression, and fuzzy inference in a unified computing framework. The developed model, intended to serve as the core of the IIASS, provides not only risk/non-risk classification but also substantiated interpretation of risk levels as normalized scores, offering managerial value to business entities. The results presented may form the basis for further software implementation of the IIASS, integration with corporate ERP systems, and functional expansion through ensemble learning strategies.
Hnatchenko et al. (Thu,) studied this question.