The problem decision of uncertainty accumulation of fuzzy-probabilistic Bayesian inference

Fuzzy Bayesian models provide ample opportunities for modeling approximate reasoning under stochastic and non-stochastic uncertainty. However, one of the main problems in using these models is to ensure the correctness and interpretability of the results of fuzzy-probabilistic Bayesian inference due to the uncertainty accumulation due to multiple consideration of the blurring of the same fuzzy variables in the repeatedly repeated fuzzy computing. The article substantiates the method of solving this problem by implementing transformations between fuzzy variables on the basis of their so-called modal interaction. The proposed method can be effectively used to model approximate reasoning based on fuzzy Bayesian models.