Abstract: Single-valued neutrosophic sets (SVN) can be used to represent the uncertainty and inconsistent information in the real situation. Information measure plays an important role in the theory of support vector network, and has attracted more and more attention in recent years. In this paper, a multi-attribute decision-making method based on single-valued neutrosophic information measure is proposed. This paper first introduces three axiomatic definitions of information metrics. It includes entropy, similarity measure and cross-entropy. Then, based on cosine function, the information measure formula is constructed, and the relationship and transformation among entropy, similarity measure and cross-entropy are discussed. On this basis, a multi-attribute decision-making method based on information measure formula is proposed. Finally, a numerical example of urban pollution assessment is given. The applicability and effectiveness of this method are demonstrated.

Key words: single-valued neutrosophic sets, entropy, information measure, cross-entropy, multi-attribute decision-making method