论文标题
您的解释有多好？算法稳定性措施，以评估深度神经网络的解释质量
How Good is your Explanation? Algorithmic Stability Measures to Assess the Quality of Explanations for Deep Neural Networks
论文作者
论文摘要
已经提出了大量方法来解释深层神经网络的决策，但是相对，几乎没有努力确保这些方法产生的解释客观上是相关的。尽管已经制定了几种可信赖解释的理想特性，但客观的措施很难得出。在这里，我们提出了两项新的措施，以评估从算法稳定性领域借来的解释：平均通用性mege和相对一致性。我们在不同的网络体系结构，常见的解释性方法和几个图像数据集上进行了广泛的实验，以证明拟议措施的好处。在与我们的公民衡量标准相比，我们不足以保证值得信赖的解释。从本质上讲，我们创建的1- lipschitz网络具有比共同的神经网络更高的含量相似的含义，而在共同的神经网络上产生了更高的解释。这表明1-Lipschitz网络是朝着更容易解释和值得信赖的预测变量的相关方向。
A plethora of methods have been proposed to explain how deep neural networks reach their decisions but comparatively, little effort has been made to ensure that the explanations produced by these methods are objectively relevant. While several desirable properties for trustworthy explanations have been formulated, objective measures have been harder to derive. Here, we propose two new measures to evaluate explanations borrowed from the field of algorithmic stability: mean generalizability MeGe and relative consistency ReCo. We conduct extensive experiments on different network architectures, common explainability methods, and several image datasets to demonstrate the benefits of the proposed measures.In comparison to ours, popular fidelity measures are not sufficient to guarantee trustworthy explanations.Finally, we found that 1-Lipschitz networks produce explanations with higher MeGe and ReCo than common neural networks while reaching similar accuracy. This suggests that 1-Lipschitz networks are a relevant direction towards predictors that are more explainable and trustworthy.
