Operational risk estimation using probabilistic model

Operational risks remain one of the most difficult and relevant issues of risk management for companies which is interested in profitable operation and sustainable development. The aim of the work is analysis of existing solutions and proposes an effective method of forecasting losses from operational risks. Suggested methods are used retrospective analysis of data. Standard methods don't take into account the causal relationships and uncertainties which are typical for individual enterprises and ineffective. The paper proposes information technology based on the adaptive approach for construction of mathematical models, in particular using Bayesian belief networks, which in turn allow processing the uncertainties caused by the random nature of risk factors.