复杂模型在风险 ALARP 情况下的评估
应用统计
2025-08-01 v2
摘要
高级计算建模的日益普及为提高安全性、效能和排放减排提供了新的机遇。将复杂模型应用于支持工程决策的进程在与其他行业相比仍较慢，这反映了不安全应用后果更大的风险。采用复杂模型会引入一种\emph{模型风险}，即错误或无帮助输出的预期后果。这应权衡更精细模型所能提供的潜在收益，同时考虑现有做法的非零风险。证明所提出机器学习应用的模型风险是否达到“尽可能低”(ALARP)水平，有助于确保在适当情况下，安全关键行业从复杂模型中受益，同时避免其滥用。本文 presented an example of automated weld radiograph classification to demonstrate how this can be achieved by combining statistical decision analysis, uncertainty quantification, and value of information.
引用
@article{arxiv.2507.10817,
title = {Is Your Model Risk ALARP? Evaluating Prospective Safety-Critical Applications of Complex Models},
author = {Domenic Di Francesco and Alan Forrest and Fiona McGarry and Nicholas Hall and Adam Sobey},
journal= {arXiv preprint arXiv:2507.10817},
year = {2025}
}