paper-with-me

Papers

Generalization Error Bound for Hyperbolic Ordinal Embedding

2021-05-21 · Atsushi Suzuki, Atsushi Nitanda, Jing Wang, Linchuan Xu, Marc Cavazza, Kenji Yamanishi

Hyperbolic ordinal embedding (HOE) represents entities as points in hyperbolic space so that they agree as well as possible with given constraints in the form of entity i is more similar to entity j than to entity k. It has been experimentally shown that HOE can obtain representations of hierarchical data such as a knowledge base and a citation network effectively, owing to hyperbolic space's exponential growth property. However, its theoretical analysis has been limited to ideal noiseless settings, and its generalization error in compensation for hyperbolic space's exponential representation ability has not been guaranteed. The difficulty is that existing generalization error bound derivations for ordinal embedding based on the Gramian matrix do not work in HOE, since hyperbolic space is not inner-product space. In this paper, through our novel characterization of HOE with decomposed Lorentz Gramian matrices, we provide a generalization error bound of HOE for the first time, which is at most exponential with respect to the embedding space's radius. Our comparison between the bounds of HOE and Euclidean ordinal embedding shows that HOE's generalization error is reasonable as a cost for its exponential representation ability.

📄 PDF Abstract BibTeX arXiv:2105.10475

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Generalization Bounds for Graph Embedding Using Negative Sampling: Linear vs Hyperbolic

2021-12-01 · NeurIPS 2021 12 · Atsushi Suzuki, Atsushi Nitanda, Jing Wang, Linchuan Xu 외

Graph embedding, which represents real-world entities in a mathematical space, has enabled numerous applications such as analyzing natural languages, social networks, biochemical networks, and knowledge bases.It has been…

Generalization BoundsGraph Embedding

Tight and fast generalization error bound of graph embedding in metric space

2023-05-13 · Atsushi Suzuki, Atsushi Nitanda, Taiji Suzuki, Jing Wang 외

Recent studies have experimentally shown that we can achieve in non-Euclidean metric space effective and efficient graph embedding, which aims to obtain the vertices' representations reflecting the graph's structure in t…

Graph Embedding

Numerically Accurate Hyperbolic Embeddings Using Tiling-Based Models

2019-12-01 · NeurIPS 2019 12 · Tao Yu, Christopher M. De Sa

Hyperbolic embeddings achieve excellent performance when embedding hierarchical data structures like synonym or type hierarchies, but they can be limited by numerical error when ordinary floating-point numbers are used t…

Finite Sample Prediction and Recovery Bounds for Ordinal Embedding

2016-06-22 · NeurIPS 2016 12 · Lalit Jain, Kevin Jamieson, Robert Nowak

The goal of ordinal embedding is to represent items as points in a low-dimensional Euclidean space given a set of constraints in the form of distance comparisons like "item $i$ is closer to item $j$ than item $k$". Ordin…

Prediction

Hyperbolic Image Segmentation

2022-03-11 · CVPR 2022 1 · Mina GhadimiAtigh, Julian Schoep, Erman Acar, Nanne van Noord 외

For image segmentation, the current standard is to perform pixel-level optimization and inference in Euclidean output embedding spaces through linear hyperplanes. In this work, we show that hyperbolic manifolds provide a…

Image SegmentationSegmentationSemantic Segmentation