arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2103.00243cs.LG
收藏

寻找鲁棒性:噪声分类任务的损失学习

Searching for Robustness: Loss Learning for Noisy Classification Tasks

University of Edinburgh(爱丁堡大学);Samsung AI Centre Cambridge(三星剑桥人工智能中心)

Boyan Gao, Henry Gouk, Timothy M. Hospedales

摘要:

我们提出了一个“学习学习”的方法,自动构建白盒分类损失函数,是强大的标签噪声的训练数据。我们使用泰勒多项式参数化一个灵活的损失函数族,并应用进化策略在这个空间中搜索噪声鲁棒损失。为了学习可应用于新任务的可重用损失函数,我们的适应度函数在一系列训练数据集和架构组合中对其性能进行了综合评分。由此产生的白盒损失提供了一个简单而快速的“即插即用”模块,该模块能够在不同的下游任务中进行有效的噪声鲁棒学习,而不需要特殊的训练过程或网络架构。我们的方法的有效性证明了各种数据集与合成和真实的标签噪声,我们比较有利,以前的工作。

英文摘要:

We present a "learning to learn" approach for automatically constructing white-box classification loss functions that are robust to label noise in the training data. We parameterize a flexible family of loss functions using Taylor polynomials, and apply evolutionary strategies to search for noise-robust losses in this space. To learn re-usable loss functions that can apply to new tasks, our fitness function scores their performance in aggregate across a range of training dataset and architecture combinations. The resulting white-box loss provides a simple and fast "plug-and-play" module that enables effective noise-robust learning in diverse downstream tasks, without requiring a special training procedure or network architecture. The efficacy of our method is demonstrated on a variety of datasets with both synthetic and real label noise, where we compare favourably to previous work.

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。

↑