arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2003.06430cs.CV
收藏

基于互信息反向传播的无偏表示学习

Learning Unbiased Representations via Mutual Information Backpropagation

Istituto Italiano di Tecnologia(意大利理工学院);University of Genova(热那亚大学);Naver Labs Europe(欧洲Naver实验室);University of Verona(维罗纳大学);Huawei Technologies Ltd.(华为技术有限公司)

Ruggero Ragonesi, Riccardo Volpi, Jacopo Cavazza, Vittorio Murino

摘要:

我们对学习数据驱动的表示感兴趣,这些表示可以很好地泛化,即使在固有偏见的数据上训练也是如此。特别是,我们面临的情况是,如果模型学习了数据的某些属性(偏差),可能会严重损害其泛化特性。我们解决这个问题,通过信息论的镜头,利用最近的研究结果的可微估计的互信息。我们提出了一种新的端到端的优化策略,同时估计和最小化学习表示和数据属性之间的互信息。当应用于标准基准测试时,我们的模型显示出与最先进的方法相当或优越的分类性能。此外,我们的方法是一般的,足以适用于"算法公平性“的问题,具有竞争力的结果。

英文摘要:

We are interested in learning data-driven representations that can generalize well, even when trained on inherently biased data. In particular, we face the case where some attributes (bias) of the data, if learned by the model, can severely compromise its generalization properties. We tackle this problem through the lens of information theory, leveraging recent findings for a differentiable estimation of mutual information. We propose a novel end-to-end optimization strategy, which simultaneously estimates and minimizes the mutual information between the learned representation and the data attributes. When applied on standard benchmarks, our model shows comparable or superior classification performance with respect to state-of-the-art approaches. Moreover, our method is general enough to be applicable to the problem of ``algorithmic fairness'', with competitive results.

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。

↑