Toward Sampling for Deep Learning Model Diagnosis
Toward Sampling for Deep Learning Model Diagnosis
复制标题
DOI：
10.1109/icde48307.2020.00201
复制
发表时间：
2020-04
期刊：
影响因子：
--
通讯作者：
Parmita Mehta;S. Portillo;M. Balazinska;Andrew J. Connolly
中科院分区：
文献类型：
--
作者：
Parmita Mehta;S. Portillo;M. Balazinska;Andrew J. Connolly
深度学习（DL）模型在许多具有高维数据（如图像、音频和文本）的领域中实现了范式改变。然而，深度神经网络的黑盒性质不仅阻碍了在医学诊断等应用中的采用，在这些应用中，可解释性是必不可少的，而且还阻碍了对表现不佳的模型的诊断。诊断或解释DL模型的任务需要计算额外的伪影，例如激活值和梯度。这些文物是大量的，他们的计算，存储和查询提出了显着的数据managementchallenges.In本文中，我们开发了一种新的数据采样技术，产生近似但准确的结果，这些模型调试查询。我们的采样技术利用DL模型学习的低维表示，并专注于此低维空间中数据的模型决策边界。
Deep learning (DL) models have achieved paradigm-changing performance in many fields with high dimensional data, such as images, audio, and text. However, the black-box nature of deep neural networks is not only a barrier to adoption in applications such as medical diagnosis, where interpretability is essential, but it also impedes diagnosis of under performing models. The task of diagnosing or explaining DL models requires the computation of additional artifacts, such as activation values and gradients. These artifacts are large in volume, and their computation, storage, and querying raise significant data management challenges.In this paper, we develop a novel data sampling technique that produces approximate but accurate results for these model debugging queries. Our sampling technique utilizes the lower dimension representation learned by the DL model and focuses on model decision boundaries for the data in this lower dimensional space.