论文标题
使用深度学习的临床风险分层模型中的不稳定性
Instability in clinical risk stratification models using deep learning
论文作者
论文摘要
尽管在ML社区中众所周知，深度学习模型遭受了不稳定性的影响，但对医疗保健部署的后果仍在特征。我们使用一组门诊预测任务作为案例研究，研究了接受电子健康记录训练的不同模型体系结构的稳定性。我们表明，即使全球性能指标保持稳定，在相同培训数据上的相同深度学习模型的重复训练可能会在患者水平上产生明显不同的结果。我们提出了两个稳定性指标，以衡量模型训练的随机性以及改善模型稳定性的缓解策略。
While it has been well known in the ML community that deep learning models suffer from instability, the consequences for healthcare deployments are under characterised. We study the stability of different model architectures trained on electronic health records, using a set of outpatient prediction tasks as a case study. We show that repeated training runs of the same deep learning model on the same training data can result in significantly different outcomes at a patient level even though global performance metrics remain stable. We propose two stability metrics for measuring the effect of randomness of model training, as well as mitigation strategies for improving model stability.
