arXivDaily arXiv每日学术速递 周一至周五更新
看起来今天可以划水了
arXiv 2102.08501cs.LGstat.ML
收藏

DEUP:直接认知不确定性预测

DEUP: Direct Epistemic Uncertainty Prediction

Mila - Quebec AI Institute(米拉-魁北克人工智能研究所);Université de Montréal(蒙特利尔大学);Massachusetts Institute of Technology(麻省理工学院)

Salem Lahlou, Moksh Jain, Hadi Nekoei, Victor Ion Butoi, Paul Bertin, Jarrid Rector-Brooks, Maksym Korablyov, Yoshua Bengio

摘要:

认知不确定性是对学习者缺乏知识的一种衡量,随着证据的增加而减少。虽然现有的工作集中在使用贝叶斯后验方差由于参数的不确定性作为衡量认知的不确定性,我们认为,这并没有捕捉到的一部分,缺乏知识引起的模型误指定。我们讨论如何超额风险,这是预测和贝叶斯预测的泛化误差之间的差距,是一个很好的措施,认识的不确定性,捕捉模型误设定的影响。因此,我们提出了一个原则性的框架,通过学习泛化误差的次级预测器并减去任意不确定性的估计,即,内在的不可预测性我们讨论了这种新的认识不确定性的措施的优点,并强调它是如何不同于基于方差的认识不确定性的措施,并解决其主要的陷阱。我们的框架,直接认知不确定性预测(DEUP)是特别有趣的互动学习环境中,学习者被允许在每一轮获得新的例子。通过大量的实验,我们说明了如何在顺序模型优化的现有方法可以改善与DEUP的认知不确定性估计,以及如何DEUP可以用来驱动探索强化学习。我们还评估了DEUP的概率图像分类和预测药物组合协同作用的不确定性估计的质量。

英文摘要:

Epistemic Uncertainty is a measure of the lack of knowledge of a learner which diminishes with more evidence. While existing work focuses on using the variance of the Bayesian posterior due to parameter uncertainty as a measure of epistemic uncertainty, we argue that this does not capture the part of lack of knowledge induced by model misspecification. We discuss how the excess risk, which is the gap between the generalization error of a predictor and the Bayes predictor, is a sound measure of epistemic uncertainty which captures the effect of model misspecification. We thus propose a principled framework for directly estimating the excess risk by learning a secondary predictor for the generalization error and subtracting an estimate of aleatoric uncertainty, i.e., intrinsic unpredictability. We discuss the merits of this novel measure of epistemic uncertainty, and highlight how it differs from variance-based measures of epistemic uncertainty and addresses its major pitfall. Our framework, Direct Epistemic Uncertainty Prediction (DEUP) is particularly interesting in interactive learning environments, where the learner is allowed to acquire novel examples in each round. Through a wide set of experiments, we illustrate how existing methods in sequential model optimization can be improved with epistemic uncertainty estimates from DEUP, and how DEUP can be used to drive exploration in reinforcement learning. We also evaluate the quality of uncertainty estimates from DEUP for probabilistic image classification and predicting synergies of drug combinations.

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。

↑