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 2003.05271cs.NEcs.NAmath.NAstat.ML
收藏

神经常微分方程中加速子波传播的插值技术

Interpolation Technique to Speed Up Gradients Propagation in Neural ODEs

Talgat Daulbaev, Alexandr Katrutsa, Larisa Markeeva, Julia Gusak, Andrzej Cichocki, Ivan Oseledets

摘要:

我们提出了一个简单的基于插值的方法,在神经ODE模型的梯度的有效近似。我们比较它与反向动态方法(在文献中称为“伴随方法”)的训练神经ODE分类,密度估计,推理近似任务。我们还提出了一个理论上的理由,我们的方法使用对数范数形式主义。因此,我们的方法允许比反向动态方法更快的模型训练,反向动态方法通过几个标准基准的大量数值实验进行了确认和验证。

英文摘要:

We propose a simple interpolation-based method for the efficient approximation of gradients in neural ODE models. We compare it with the reverse dynamic method (known in the literature as "adjoint method") to train neural ODEs on classification, density estimation, and inference approximation tasks. We also propose a theoretical justification of our approach using logarithmic norm formalism. As a result, our method allows faster model training than the reverse dynamic method that was confirmed and validated by extensive numerical experiments for several standard benchmarks.

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。

↑