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 2403.02473cs.CV
收藏

卷积神经网络何时停止学习?

When do Convolutional Neural Networks Stop Learning?

College of Business, Texas A&M University(德州农工大学商学院);University of Louisiana at Lafayette(路易斯安那大学拉法叶分校)

Sahan Ahmad, Gabriel Trahan, Aminul Islam

AI总结:

本研究提出一种基于CNN各层数据变化预估其接近最优学习容量的假设,无需验证数据即可提前停止训练,在通用与医学图像数据集上均能在不损失精度的前提下显著节省计算时间。

摘要:

卷积神经网络(CNNs)已在图像分类、检测、分割及医学图像分析等计算机视觉任务中展现出卓越性能。通常,人们会设定任意数量的epoch来训练此类神经网络。在单个epoch内,全部训练数据(按批量大小划分)会被馈送入网络。实践中,通常结合验证误差与训练损失来评估神经网络的泛化能力,这一指标反映了网络的最优学习容量。当前的常规做法是,当训练损失下降、而训练误差与验证误差之间的差距(即泛化差距)增大时停止训练,以避免过拟合。然而,这是一种基于试错的方法,由此引出一个关键问题:能否仅基于训练数据预估神经网络何时停止学习?\n本研究提出了一项假设,通过分析CNN变体各层间的数据变化,来预测其接近最优的学习容量。在训练阶段,我们利用该假设无需任何验证数据即可预估CNN变体的接近最优学习容量。该假设可即插即用地部署于任何现有CNN变体,且不会为网络引入额外可训练参数。我们在6种不同的CNN变体以及3个通用图像数据集(CIFAR10、CIFAR100和SVHN)上验证了该假设。基于这些CNN变体和数据集的结果显示,该假设平均可节省58.49%的训练计算时间。我们进一步在10个医学图像数据集上验证了该假设,并与MedMNIST-V2基准进行对比。实验结果表明,相较于MedMNIST-V2基准,我们在不损失精度的前提下节省了约44.1%的计算时间。

英文摘要:

Convolutional Neural Networks (CNNs) have demonstrated outstanding performance in computer vision tasks such as image classification, detection, segmentation, and medical image analysis. In general, an arbitrary number of epochs is used to train such neural networks. In a single epoch, the entire training data -- divided by batch size -- are fed to the network. In practice, validation error with training loss is used to estimate the neural network's generalization, which indicates the optimal learning capacity of the network. Current practice is to stop training when the training loss decreases and the gap between training and validation error increases (i.e., the generalization gap) to avoid overfitting. However, this is a trial-and-error-based approach which raises a critical question: Is it possible to estimate when neural networks stop learning based on training data? This research work introduces a hypothesis that analyzes the data variation across all the layers of a CNN variant to anticipate its near-optimal learning capacity. In the training phase, we use our hypothesis to anticipate the near-optimal learning capacity of a CNN variant without using any validation data. Our hypothesis can be deployed as a plug-and-play to any existing CNN variant without introducing additional trainable parameters to the network. We test our hypothesis on six different CNN variants and three different general image datasets (CIFAR10, CIFAR100, and SVHN). The result based on these CNN variants and datasets shows that our hypothesis saves 58.49\% of computational time (on average) in training. We further conduct our hypothesis on ten medical image datasets and compared with the MedMNIST-V2 benchmark. Based on our experimental result, we save $\approx$ 44.1\% of computational time without losing accuracy against the MedMNIST-V2 benchmark.

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。

↑