arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2210.04686cs.LGcs.AIeess.SP
收藏

利用可解释的AI提高神经网络的性能

Utilizing Explainable AI for improving the Performance of Neural Networks

Infineon Technologies AG(英飞凌科技股份公司);Friedrich-Alexander-University Erlangen-Nuremberg(埃尔朗根-纽伦堡弗里德里希-亚历山大大学);Technical University of Munich(慕尼黑工业大学)

Huawei Sun, Lorenzo Servadei, Hao Feng, Michael Stephan, Robert Wille, Avik Santra

摘要:

如今,深度神经网络被广泛应用于对社会有直接影响的各种领域。尽管这些模型通常表现出出色的性能,但它们长期以来一直被用作黑匣子。为了解决这个问题,可解释的人工智能(XAI)一直在发展成为一个旨在提高模型透明度并增加其可信度的领域。我们提出了一个再训练管道,从XAI开始并利用最先进的技术不断改进模型预测。为此,我们使用XAI结果,即SHapley加法解释(SHAP)值,为数据样本提供特定的训练权重。这导致改进的模型训练,从而提高性能。为了对我们的方法进行基准测试,我们在现实生活和公共数据集上对其进行了评估。首先,我们在基于雷达的人员计数场景上执行该方法。之后,我们在CIFAR-10(一个公共计算机视觉数据集)上进行了测试。实验使用基于形状的再训练方法实现了4%以上的准确性w.r.t.对人数统计任务进行标准的等权重再培训。此外,在CIFAR-10上,我们基于SHAP的加权策略最终比具有相等加权样本的训练过程具有3%的准确率。

英文摘要:

Nowadays, deep neural networks are widely used in a variety of fields that have a direct impact on society. Although those models typically show outstanding performance, they have been used for a long time as black boxes. To address this, Explainable Artificial Intelligence (XAI) has been developing as a field that aims to improve the transparency of the model and increase their trustworthiness. We propose a retraining pipeline that consistently improves the model predictions starting from XAI and utilizing state-of-the-art techniques. To do that, we use the XAI results, namely SHapley Additive exPlanations (SHAP) values, to give specific training weights to the data samples. This leads to an improved training of the model and, consequently, better performance. In order to benchmark our method, we evaluate it on both real-life and public datasets. First, we perform the method on a radar-based people counting scenario. Afterward, we test it on the CIFAR-10, a public Computer Vision dataset. Experiments using the SHAP-based retraining approach achieve a 4% more accuracy w.r.t. the standard equal weight retraining for people counting tasks. Moreover, on the CIFAR-10, our SHAP-based weighting strategy ends up with a 3% accuracy rate than the training procedure with equal weighted samples.

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。

↑