arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2104.06323cs.LGcs.AIstat.ML
收藏

δ- CLUE:不确定性估计的不同解释集

δ-CLUE: Diverse Sets of Explanations for Uncertainty Estimates

University of Cambridge(剑桥大学);The Alan Turing Institute(艾伦图灵研究所)

Dan Ley, Umang Bhatt, Adrian Weller

摘要:

为了解释可微概率模型的不确定性估计,最近的工作提出了生成反事实潜在不确定性假设(CLUE)。然而,对于单个输入,由于缺乏对解释的约束,这种方法可以输出各种解释。在这里,我们增强了原始的CLUE方法,以提供我们称之为$δ$ -CLUE的方法。CLUE指示$\it{one}$改变输入的方式,同时保留在数据流形上,使得模型对其预测变得更有信心。相反,我们返回一个$\it{set}$的合理线索:多个不同的输入,在潜在空间中的原始输入的$δ$球内,所有这些都产生可信的预测。

英文摘要:

To interpret uncertainty estimates from differentiable probabilistic models, recent work has proposed generating Counterfactual Latent Uncertainty Explanations (CLUEs). However, for a single input, such approaches could output a variety of explanations due to the lack of constraints placed on the explanation. Here we augment the original CLUE approach, to provide what we call $δ$-CLUE. CLUE indicates $\it{one}$ way to change an input, while remaining on the data manifold, such that the model becomes more confident about its prediction. We instead return a $\it{set}$ of plausible CLUEs: multiple, diverse inputs that are within a $δ$ ball of the original input in latent space, all yielding confident predictions.

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。

↑