arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2005.03847cs.LGmath.MGstat.ML
收藏

树!我不是树!我是一个低维双曲嵌入

Tree! I am no Tree! I am a Low Dimensional Hyperbolic Embedding

University of Michigan(密歇根大学);Yale University(耶鲁大学);Department of Mathematics(数学系);Department of Statistics and Data Science(统计与数据科学系);Ann Arbor, MI, 48104(安阿伯,密歇根州,48104);New Haven, CT, 06510(纽黑文,康涅狄格州,06510)

Rishi Sonthalia, Anna C. Gilbert

摘要:

给定数据,找到数据的忠实低维双曲嵌入是提取数据层次信息或学习数据代表性几何特征的关键方法。在本文中,我们探索了一种新的方法学习双曲表示采取度量优先的方法。我们不是直接确定低维双曲嵌入,而是学习数据上的树结构。然后,该树结构可以直接用于提取分层信息,使用Sarkar的构造\cite{sarkar}嵌入到双曲流形中,或者用作原始度量的树近似。为此,我们提出了一种新的快速算法\textsc{TreeRep},这样,给定一个$δ$ -双曲度量(对于任何$δ\geq 0$),该算法学习一个树结构,近似原始度量。在$δ= 0$的情况下,我们分析表明,\textsc{TreeRep}完全恢复原来的树结构。我们的经验表明,\textsc{TreeRep}不仅是许多数量级的速度比以前已知的算法,但也产生的度量较低的平均失真和较高的平均精度比以前的算法学习双曲嵌入,提取层次信息,并通过树度量近似度量。

英文摘要:

Given data, finding a faithful low-dimensional hyperbolic embedding of the data is a key method by which we can extract hierarchical information or learn representative geometric features of the data. In this paper, we explore a new method for learning hyperbolic representations by taking a metric-first approach. Rather than determining the low-dimensional hyperbolic embedding directly, we learn a tree structure on the data. This tree structure can then be used directly to extract hierarchical information, embedded into a hyperbolic manifold using Sarkar's construction \cite{sarkar}, or used as a tree approximation of the original metric. To this end, we present a novel fast algorithm \textsc{TreeRep} such that, given a $δ$-hyperbolic metric (for any $δ\geq 0$), the algorithm learns a tree structure that approximates the original metric. In the case when $δ= 0$, we show analytically that \textsc{TreeRep} exactly recovers the original tree structure. We show empirically that \textsc{TreeRep} is not only many orders of magnitude faster than previously known algorithms, but also produces metrics with lower average distortion and higher mean average precision than most previous algorithms for learning hyperbolic embeddings, extracting hierarchical information, and approximating metrics via tree metrics.

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。

↑