IEEE Transactions on Fuzzy Systems | 2019
A Knowledge-Based Risk Measure From the Fuzzy Multicriteria Decision-Making Perspective
Abstract
Risk measures play significant roles in determining the magnitude of risks. The traditional risk measures consider only the consequence <inline-formula><tex-math notation= LaTeX >$(C)$</tex-math></inline-formula> and the probability <inline-formula><tex-math notation= LaTeX >$(P)$</tex-math></inline-formula> and ignore the support of the knowledge behind to estimate <inline-formula><tex-math notation= LaTeX >$C$</tex-math></inline-formula> and <inline-formula><tex-math notation= LaTeX >$P$</tex-math></inline-formula>. Several researchers have suggested adding knowledge as a third dimension in the risk measures. However, the issues of how to embed the dimension of knowledge in the risk measures to output an explicit expression of the risk measure and how to measure the strength of knowledge remain unresolved. This paper proposes a new risk measure incorporating the dimension of knowledge, apart from <inline-formula><tex-math notation= LaTeX >$C$</tex-math></inline-formula> and <inline-formula><tex-math notation= LaTeX >$P$</tex-math></inline-formula>. It is shown that the proposed risk measure has the form of traditional risk measures when the risk assessor has full knowledge. In addition, a fuzzy multicriteria decision-making (MCDM) method is employed to assess the strength of knowledge. In the fuzzy MCDM method, an entropy optimization problem is solved to obtain fuzzy measures, which are critical for determining the score of the strength of knowledge. Finally, the proposed method is applied to a project risk assessment, showing the feasibility of the method.