arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2106.07875stat.MLcs.LG
收藏

S-LIME:用于模型说明的稳定石灰

S-LIME: Stabilized-LIME for Model Explanation

Cornell University(康奈尔大学);Weill Cornell Medicine(威尔康奈尔医学院)

Zhengze Zhou, Giles Hooker, Fei Wang

摘要:

越来越多的机器学习模型被部署在金融和医疗保健等高风险领域。尽管它们的性能优越,但许多模型本质上是难以解释的黑匣子。研究人员正在努力开发解释这些黑箱模型的方法。基于扰动的事后解释(如LIME)是在机器学习模型构建后对其进行解释的广泛使用的方法。这类方法已被证明表现出很大的不稳定性,对方法本身的有效性构成严重挑战,并损害用户的信任。在本文中,我们提出了S-LIME,它利用一个假设检验框架的基础上中心极限定理确定扰动点的数量,以保证稳定的解释。仿真和真实世界的数据集上的实验证明了我们的方法的有效性。

英文摘要:

An increasing number of machine learning models have been deployed in domains with high stakes such as finance and healthcare. Despite their superior performances, many models are black boxes in nature which are hard to explain. There are growing efforts for researchers to develop methods to interpret these black-box models. Post hoc explanations based on perturbations, such as LIME, are widely used approaches to interpret a machine learning model after it has been built. This class of methods has been shown to exhibit large instability, posing serious challenges to the effectiveness of the method itself and harming user trust. In this paper, we propose S-LIME, which utilizes a hypothesis testing framework based on central limit theorem for determining the number of perturbation points needed to guarantee stability of the resulting explanation. Experiments on both simulated and real world data sets are provided to demonstrate the effectiveness of our method.

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。

↑