arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1609.08017cs.LGstat.ML
收藏

期望线性正则化的Dropout

Dropout with Expectation-linear Regularization

Carnegie Mellon University(卡内基梅隆大学);Harvard University(哈佛大学)

Xuezhe Ma, Yingkai Gao, Zhiting Hu, Yaoliang Yu, Yuntian Deng, Eduard Hovy

摘要:

Dropout是一种简单而有效的训练深度神经网络的方法,已经取得了许多令人印象深刻的经验成功,并催生了许多最近的理论研究。然而,由于易处理性的考虑,dropout的训练和推理阶段之间的差距在很大程度上仍然没有得到充分的重视。在这项工作中,我们首先制定辍学作为一个易于处理的近似一些潜在的变量模型,导致一个干净的参数共享视图,并使进一步的理论分析。然后,我们引入(近似)期望线性丢弃神经网络,其推理间隙,我们能够正式表征。在数学上,我们表明,我们提出的测量的推理间隙可以用来规范的标准辍学的训练目标,从而在\emph{explicit}控制的差距。我们的方法是简单和有效的标准dropout。我们进一步证明了由于期望线性化的精度损失的上界,描述类的输入分布,期望线性化容易。在三个图像分类基准数据集上的实验表明,减少推理间隙确实可以一致地提高性能。

英文摘要:

Dropout, a simple and effective way to train deep neural networks, has led to a number of impressive empirical successes and spawned many recent theoretical investigations. However, the gap between dropout's training and inference phases, introduced due to tractability considerations, has largely remained under-appreciated. In this work, we first formulate dropout as a tractable approximation of some latent variable model, leading to a clean view of parameter sharing and enabling further theoretical analysis. Then, we introduce (approximate) expectation-linear dropout neural networks, whose inference gap we are able to formally characterize. Algorithmically, we show that our proposed measure of the inference gap can be used to regularize the standard dropout training objective, resulting in an \emph{explicit} control of the gap. Our method is as simple and efficient as standard dropout. We further prove the upper bounds on the loss in accuracy due to expectation-linearization, describe classes of input distributions that expectation-linearize easily. Experiments on three image classification benchmark datasets demonstrate that reducing the inference gap can indeed improve the performance consistently.

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。

↑