Answers are grounded in this paper and cite their sources.
There is no common consensus on human-AI synergistic risk evaluating framework among the extant studies, especially when simultaneously addressing the challenges caused by potential low-probability incident risks and highly sensitive quality-deterioration nature in perishable-food manufacturing and logistics services. We view AI-based technology as an effective incident-scenario forecasting tool when facing external environmental threats. We use a well-trained AdaBoost model to forecast the occurrence probability of each potential incident by observing and inputting real-time data about the nine key influencing factors. From the internal quality-deterioration perspective, we advocate human-expertise-dominated food-safety posterior information revision. In this specific framework, we therefore employ a hybrid approach by incorporating AdaBoost, cumulative prospect theory-based hesitant Pythagorean fuzzy set, and Graphic Evaluation and Review Technique to address the comprehensive risk-evaluating issues of scenario forecasting, response timing, and spatial ranking. The main findings reveal that our approach not only has advantages in mitigating the mixed ambiguity of internal quality deterioration and external environmental threats, but also has good performance and reliable robustness in a changing parameter environment.
No takes yet. Share an insight, caveat, or question.
Zhan et al. (2025) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: