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 2006.08210cs.LGstat.ML
收藏

双曲神经网络++

Hyperbolic Neural Networks++

RIKEN AIP(理化学研究所人工智能项目);The University of Tokyo(东京大学)

Ryohei Shimizu, Yusuke Mukuta, Tatsuya Harada

摘要:

双曲空间由于其指数体积增长而具有嵌入树结构而不失真的能力,最近已被应用于机器学习以更好地捕获数据的分层性质。在这项研究中,我们推广的基本组成部分的神经网络在一个单一的双曲几何模型,即庞加莱球模型。这种新颖的方法在统一的数学解释下构建了多项式逻辑回归,全连接层,卷积层和注意力机制,而不增加参数。实验结果表明,我们的方法相比,传统的双曲分量,稳定性和优于他们的欧几里德同行的参数效率。

英文摘要:

Hyperbolic spaces, which have the capacity to embed tree structures without distortion owing to their exponential volume growth, have recently been applied to machine learning to better capture the hierarchical nature of data. In this study, we generalize the fundamental components of neural networks in a single hyperbolic geometry model, namely, the Poincaré ball model. This novel methodology constructs a multinomial logistic regression, fully-connected layers, convolutional layers, and attention mechanisms under a unified mathematical interpretation, without increasing the parameters. Experiments show the superior parameter efficiency of our methods compared to conventional hyperbolic components, and stability and outperformance over their Euclidean counterparts.

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。

↑