arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2107.13625cs.LGcs.AIcs.CVcs.CY
收藏

未知变异敏感因子的适应与推广

Adaptation and Generalization for Unknown Sensitive Factors of Variations

Johns Hopkins University Applied Physics Laboratory(约翰斯·霍普金斯大学应用物理实验室)

William Paul, Philippe Burlina

摘要:

在不受限制的设置中确保AI是一个关键问题。我们的框架解决了人工智能保证的挑战,这些挑战位于领域自适应、公平性和反事实分析的交叉点,通过发现和干预显著影响人工智能模型鲁棒性的数据变化因素(例如天气或照明条件)来进行操作。稳健性在这里被理解为模型性能对敏感因子变化的不敏感性。敏感因素传统上是在监督环境中设定的,其中因素是先验已知的(例如,为了公平起见,这可能是性别或种族等因素)。相比之下,我们的动机是现实生活中的场景,在这些场景中,对导致模型失败的某些因素实际上知之甚少或一无所知。这导致我们考虑各种设置(无监督,领域泛化,半监督),对应于不同程度的不完整的知识,这些因素。因此,我们的两步方法的工作原理是:a)发现导致人工智能系统以无监督方式失败的敏感因素,然后b)干预模型以减少这些因素的影响。我们的方法考虑了3种干预措施,包括增强,一致性和对抗性干预(ACAI)。我们展示了对发现/源因素进行干预以推广到目标/真实因素的能力。我们还演示了如何适应真正的变化因素,可以在半监督的情况下,一些目标因子标签是已知的,通过自动干预选择。实验表明,我们的方法在实现最佳效用与灵敏度/鲁棒性权衡方面比基线模型有所改进。

英文摘要:

Assured AI in unrestricted settings is a critical problem. Our framework addresses AI assurance challenges lying at the intersection of domain adaptation, fairness, and counterfactuals analysis, operating via the discovery and intervention on factors of variations in data (e.g. weather or illumination conditions) that significantly affect the robustness of AI models. Robustness is understood here as insensitivity of the model performance to variations in sensitive factors. Sensitive factors are traditionally set in a supervised setting, whereby factors are known a-priori (e.g. for fairness this could be factors like sex or race). In contrast, our motivation is real-life scenarios where less, or nothing, is actually known a-priori about certain factors that cause models to fail. This leads us to consider various settings (unsupervised, domain generalization, semi-supervised) that correspond to different degrees of incomplete knowledge about those factors. Therefore, our two step approach works by a) discovering sensitive factors that cause AI systems to fail in a unsupervised fashion, and then b) intervening models to lessen these factor's influence. Our method considers 3 interventions consisting of Augmentation, Coherence, and Adversarial Interventions (ACAI). We demonstrate the ability for interventions on discovered/source factors to generalize to target/real factors. We also demonstrate how adaptation to real factors of variations can be performed in the semi-supervised case where some target factor labels are known, via automated intervention selection. Experiments show that our approach improves on baseline models, with regard to achieving optimal utility vs. sensitivity/robustness tradeoffs.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑