On stability of fuzzy formal concepts over randomized one-sided formal context

摘要

AbstractWe propose a probabilistic approach to the issue of one-sided fuzzy formal concepts stability. The modified Rice–Siff algorithm represents a crisp index how to select the relevant concepts from the set of all one-sided fuzzy formal concepts. We suggest to explore the formal concepts stability affected by the random fluctuation of values in a formal context. We describe the algorithm and study the properties of the concept stability using random variables with the Gaussian normal distribution. In combination with the modified Rice–Siff algorithm, the Gaussian probabilistic index improves the analysis of the most relevant one-sided formal concepts from the original one-sided formal context. The connections to recent works in the related directions are presented.