arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2107.06217cs.LGcs.AI
收藏

什么分类器知道他们不知道的?

What classifiers know what they don't?

Facebook AI Research(脸书人工智能研究院)

Mohamed Ishmael Belghazi, David Lopez-Paz

摘要:

面对未知时的不确定性是智能决策的关键。然而,机器学习算法缺乏关于其预测不确定性的可靠估计。这会导致在训练过程中遇到看不见的类时做出错误和过于自信的决定。尽管为分类器配备为现实世界准备的不确定性估计很重要,但之前的工作主要集中在小数据集上,并且训练和测试数据之间几乎没有类别差异。为了缩小这一差距,我们引入了UIMNET:一个现实的、ImageNet规模的测试平台,用于评估深度图像分类器的预测不确定性估计。我们的基准测试提供了八种最先进的算法,六种不确定性度量,四种域内度量,三种域外度量的实现,以及一个全自动的管道来训练,校准,集成,选择和评估模型。我们的测试台是开源的,我们所有的结果都可以从我们存储库中的固定提交中重现。添加新的数据集、算法、度量或指标只需要几行代码,因此希望UIMNET成为不确定性估计中现实、严格和可重复研究的垫脚石。我们的研究结果表明,合奏的ERM分类器,以及单一的MIMO分类器是两个最好的替代品,目前可用于测量的不确定性域和域外类。

英文摘要:

Being uncertain when facing the unknown is key to intelligent decision making. However, machine learning algorithms lack reliable estimates about their predictive uncertainty. This leads to wrong and overly-confident decisions when encountering classes unseen during training. Despite the importance of equipping classifiers with uncertainty estimates ready for the real world, prior work has focused on small datasets and little or no class discrepancy between training and testing data. To close this gap, we introduce UIMNET: a realistic, ImageNet-scale test-bed to evaluate predictive uncertainty estimates for deep image classifiers. Our benchmark provides implementations of eight state-of-the-art algorithms, six uncertainty measures, four in-domain metrics, three out-domain metrics, and a fully automated pipeline to train, calibrate, ensemble, select, and evaluate models. Our test-bed is open-source and all of our results are reproducible from a fixed commit in our repository. Adding new datasets, algorithms, measures, or metrics is a matter of a few lines of code-in so hoping that UIMNET becomes a stepping stone towards realistic, rigorous, and reproducible research in uncertainty estimation. Our results show that ensembles of ERM classifiers as well as single MIMO classifiers are the two best alternatives currently available to measure uncertainty about both in-domain and out-domain classes.

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。

↑