arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2305.18552cs.LGcs.NE
收藏

在神经网络中学习线性群

Learning Linear Groups in Neural Networks

Harvard University(哈佛大学);Amazon Web Services(亚马逊云科技)

Emmanouil Theodosis, Karim Helwani, Demba Ba

AI总结:

本文提出线性群网络(LGNs),无需先验对称性知识即可从数据中学习作用于权重空间的线性群,并证明其结构依赖于数据分布和下游任务。

摘要:

在神经网络中采用等变性可以通过在架构中编码领域知识来提高参数效率并改善泛化性能;然而,现有的大多数方法都需要预先指定所需的对称性。我们提出了一种神经网络架构,称为线性群网络(LGNs),用于学习作用于神经网络权重空间的线性群。线性群因其固有的可解释性而具有优势,因为它们可以表示为有限矩阵。LGNs 在没有任何监督或对数据中隐藏对称性的先验知识的情况下学习群,并且这些群可以映射到机器学习中众所周知的操作。我们使用 LGNs 在多个数据集上学习群,同时考虑不同的下游任务;我们证明线性群结构既依赖于数据分布,也依赖于所考虑的任务。

英文摘要:

Employing equivariance in neural networks leads to greater parameter efficiency and improved generalization performance through the encoding of domain knowledge in the architecture; however, the majority of existing approaches require an a priori specification of the desired symmetries. We present a neural network architecture, Linear Group Networks (LGNs), for learning linear groups acting on the weight space of neural networks. Linear groups are desirable due to their inherent interpretability, as they can be represented as finite matrices. LGNs learn groups without any supervision or knowledge of the hidden symmetries in the data and the groups can be mapped to well known operations in machine learning. We use LGNs to learn groups on multiple datasets while considering different downstream tasks; we demonstrate that the linear group structure depends on both the data distribution and the considered task.

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。

↑