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


Current browse context:


Change to browse by:

References & Citations

DBLP - CS Bibliography


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

Computer Science > Machine Learning

Title: Fast Variational AutoEncoder with Inverted Multi-Index for Collaborative Filtering

Abstract: Variational AutoEncoder (VAE) has been extended as a representative nonlinear method for collaborative filtering. However, the bottleneck of VAE lies in the softmax computation over all items, such that it takes linear costs in the number of items to compute the loss and gradient for optimization. This hinders the practical use due to millions of items in real-world scenarios. Importance sampling is an effective approximation method, based on which the sampled softmax has been derived. However, existing methods usually exploit the uniform or popularity sampler as proposal distributions, leading to a large bias of gradient estimation. To this end, we propose to decompose the inner-product-based softmax probability based on the inverted multi-index, leading to sublinear-time and highly accurate sampling. Based on the proposed proposals, we develop a fast Variational AutoEncoder (FastVAE) for collaborative filtering. FastVAE can outperform the state-of-the-art baselines in terms of both sampling quality and efficiency according to the experiments on three real-world datasets.
Comments: 11 pages, WWW 2022
Subjects: Machine Learning (cs.LG); Information Retrieval (cs.IR)
DOI: 10.1145/3485447.3512068
Cite as: arXiv:2109.05773 [cs.LG]
  (or arXiv:2109.05773v2 [cs.LG] for this version)

Submission history

From: Jin Chen [view email]
[v1] Mon, 13 Sep 2021 08:31:59 GMT (403kb,D)
[v2] Mon, 30 May 2022 07:45:34 GMT (3995kb,D)

Link back to: arXiv, form interface, contact.