Survey of Various Fuzzy and Uncertain Decision-Making Methods

Researchers Takaaki Fujita and Florentin Smarandache have released a major academic survey, 'Survey of Various Fuzzy and Uncertain Decision-Making Methods,' published as a 446-page book by the Neutrosophic Science International Association (NSIA). The work addresses a core challenge in applied AI: how to build systems that make reliable decisions with vague, incomplete, or conflicting data. It provides a structured, task-oriented taxonomy to navigate the complex field of uncertainty-aware multi-criteria decision-making (MCDM), which is critical for real-world applications in finance, engineering, and autonomous systems.

The survey systematically breaks down the discipline into key components. It first organizes problem-level settings, such as group consensus, dynamic, or multi-agent scenarios. It then details methods for weight elicitation—determining the importance of different decision factors—under fuzzy or linguistic inputs. Finally, it contrasts major solution procedures, including compensatory scoring, distance-to-reference approaches, and non-compensatory outranking frameworks. A significant contribution is its practical guidance, helping developers choose methods based on three critical axes: the required robustness of the decision, the need for interpretable rules, and the availability of data.

The authors conclude by outlining open research directions, pushing the field toward more explainable uncertainty integration and scalable methods for large-scale, dynamic environments. This survey serves as both a foundational reference for newcomers and a structured map of the landscape for experts, aiming to bridge the gap between theoretical decision models and their practical implementation in AI systems.