We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.LG

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Machine Learning

Title: LoNe Sampler: Graph node embeddings by coordinated local neighborhood sampling

Abstract: Local graph neighborhood sampling is a fundamental computational problem that is at the heart of algorithms for node representation learning. Several works have presented algorithms for learning discrete node embeddings where graph nodes are represented by discrete features such as attributes of neighborhood nodes. Discrete embeddings offer several advantages compared to continuous word2vec-like node embeddings: ease of computation, scalability, and interpretability. We present LoNe Sampler, a suite of algorithms for generating discrete node embeddings by Local Neighborhood Sampling, and address two shortcomings of previous work. First, our algorithms have rigorously understood theoretical properties. Second, we show how to generate approximate explicit vector maps that avoid the expensive computation of a Gram matrix for the training of a kernel model. Experiments on benchmark datasets confirm the theoretical findings and demonstrate the advantages of the proposed methods.
Comments: Accepted to AAAI 2023. arXiv admin note: substantial text overlap with arXiv:2102.04770
Subjects: Machine Learning (cs.LG); Data Structures and Algorithms (cs.DS)
Cite as: arXiv:2211.15114 [cs.LG]
  (or arXiv:2211.15114v1 [cs.LG] for this version)

Submission history

From: Konstantin Kutzkov [view email]
[v1] Mon, 28 Nov 2022 08:04:26 GMT (3415kb,D)

Link back to: arXiv, form interface, contact.