提出新型神经网络解释方法，避免除零问题并提升稳定性。
Relative Layer-Wise Relevance Propagation: a more Robust Neural Networks eXplaination
- 基于相对归因思想，改进传统LRP的传播机制
- 在多个图像分类模型上验证，结果更稳定可靠
- 无需调参，适用于CNN与残差网络，适合可解释性研究
机器学习在众多任务中表现优异，但缺乏决策过程的可解释性。为此，层间归因传播（LRP）是可解释机器学习的重要方法之一，旨在量化神经网络输出对输入的贡献。现有方法主要受限于小值除法带来的数值不稳定性。本文提出相对层间归因传播（R-LRP），在保持类守恒律的基础上，仅在ResNet跳跃连接处涉及小值除法，显著提升鲁棒性。该方法适用于密集连接、CNN及残差网络，在图像分类任务中可可视化像素对预测的贡献，帮助识别关键区域。实验在CIFAR-10、ImageNet等数据集上对比了简单CNN、VGG16、VGG19和ResNet50，结果表明R-LRP具有更强的稳定性且无需调参。
原文摘要 · Abstract (English)
Machine learning methods are solving very successfully a plethora of tasks, but they have the disadvantage of not providing any information about their decision. Consequently, estimating the reasoning of the system provides additional information. For this, Layer-Wise Relevance Propagation (LRP) is one of the methods in eXplainable Machine Learning (XML). Its purpose is to provide contributions of any neural network output in the domain of its input. The main drawback of current methods is mainly due to division by small values. To overcome this problem, we provide a new definition called Relative LRP where the classical conservation law is satisfied up to a multiplicative factor but without divisions by small values except for Resnet skip connection. In this article, we will focus on image classification. This allows us to visualize the contributions of a pixel to the predictions of a multi-layer neural network. Pixel contributions provide a focus to further analysis on regions of potential interest. R-LRP can be applied for any dense, CNN or residual neural networks. Moreover, R-LRP doesn't need any hyperparameters to tune contrary to other LRP methods. We then compare the R-LRP method on different datasets with simple CNN, VGG16, VGG19 and Resnet50 networks.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。