Uncertainty Merging with Basic Uncertain Information in Probability Environment
LeSheng Jin,Yi Yang,Zhen-Song Chen,Muhammet Deveci,Radko Mesiar
DOI: https://doi.org/10.1016/j.fss.2024.109153
IF: 4.462
2025-01-01
Fuzzy Sets and Systems
Abstract:Basic uncertain information is a recently introduced and significant type of uncertainty that proves particularly valuable in decision-making environments with inherent uncertainties. In this study, we propose the concept of uncertainty cognition merging, which effectively combines basic uncertain information granules with probability measures to generate new probability measures within the same probability space. Additionally, we present a degenerated method that merges basic uncertain information granules with unit intervals to create new subintervals. We introduce four distinct uncertainty cognition merging methods and thoroughly compare and analyze their respective properties, limitations, and advantages. To demonstrate the practical application potential of our proposals, we provide numerical examples alongside further mathematical results. ence information that can help them adjust their initial evaluation or prediction values. These external agencies often have access to statistical and probabilistic information based on history or big data, which generally gives them certain advantages such as appearing more objective and closer to reality and the market. However, there are also obvious drawbacks: these data are more general and may not be well-suited for specific decision-making scenarios; moreover, these data tend to be delayed, increasing the error rate of decisions. On the other hand, the core decision-making group within an organization is usually composed of internal experts or specially invited professionals who provide evaluations and predictions that are more specialized and relevant to the company's products. The downside of expert evaluations is that they tend to be subjective and with a significant amount of uncertainty. In summary, both internal evaluations and external consultations are important means of obtaining evaluation or prediction data. If internal experts assess that there is no uncertainty, decision-makers often tend to forgo seeking external consulting firms, even though these assessments may have some subjectivity, due to the high consultancy fees charged by such firms. However, in many cases, constrained by factors such as individual knowledge, experience, abilities, time and dedication of each expert, expert assessments carry significant uncertainties. This necessitates the reliance on statistical or probabilistic information provided by external consulting firms in order to make more informed assessments and decisions. The integration and merging of this external consulting information with the initial uncertain evaluations made by internal experts to obtain a more reasonable assessment result is the issue that will be discussed in this article.