Abstract
Fuzzy sets possess remarkable abilities in expressing and handling information uncertainty, which has resulted in their widespread application in various fields. Nevertheless, distance measurement between IFSs for quantitating their differences and levels of differentiation has remained an open problem that deserves attention. Despite the development of various metrics, they either lack intuitive insight or do not satisfy the axioms of distance measurement, leading to counterintuitive results. To address these issues, this paper proposed a distance measurement method based on Clark divergence, which satisfies the distance measurement axioms and exhibits nonlinearity. Numerical examples demonstrate that our method effectively distinguishes different indicators, yielding more reasonable results. Moreover, when comparing relative differences of the results, our method demonstrated superior adaptability to complex environmental decision-making, providing decision-makers with more accurate and confidential judgments. The pattern classification algorithm designed in this paper will offer a promising solution to inference problems.