Matches in SemOpenAlex for { <https://semopenalex.org/work/W2952208216> ?p ?o ?g. }
Showing items 1 to 83 of
83
with 100 items per page.
- W2952208216 abstract "Latent Dirichlet Allocation (LDA) is a popular tool for analyzing discrete count data such as text and images. Applications require LDA to handle both large datasets and a large number of topics. Though distributed CPU systems have been used, GPU-based systems have emerged as a promising alternative because of the high computational power and memory bandwidth of GPUs. However, existing GPU-based LDA systems cannot support a large number of topics because they use algorithms on dense data structures whose time and space complexity is linear to the number of topics. In this paper, we propose SaberLDA, a GPU-based LDA system that implements a sparsity-aware algorithm to achieve sublinear time complexity and scales well to learn a large number of topics. To address the challenges introduced by sparsity, we propose a novel data layout, a new warp-based sampling kernel, and an efficient sparse count matrix updating algorithm that improves locality, makes efficient utilization of GPU warps, and reduces memory consumption. Experiments show that SaberLDA can learn from billions-token-scale data with up to 10,000 topics, which is almost two orders of magnitude larger than that of the previous GPU-based systems. With a single GPU card, SaberLDA is able to learn 10,000 topics from a dataset of billions of tokens in a few hours, which is only achievable with clusters with tens of machines before." @default.
- W2952208216 created "2019-06-27" @default.
- W2952208216 creator A5003713785 @default.
- W2952208216 creator A5020580195 @default.
- W2952208216 creator A5069169048 @default.
- W2952208216 creator A5087673923 @default.
- W2952208216 date "2016-10-08" @default.
- W2952208216 modified "2023-09-25" @default.
- W2952208216 title "SaberLDA: Sparsity-Aware Learning of Topic Models on GPUs" @default.
- W2952208216 doi "https://doi.org/10.48550/arxiv.1610.02496" @default.
- W2952208216 hasPublicationYear "2016" @default.
- W2952208216 type Work @default.
- W2952208216 sameAs 2952208216 @default.
- W2952208216 citedByCount "0" @default.
- W2952208216 crossrefType "posted-content" @default.
- W2952208216 hasAuthorship W2952208216A5003713785 @default.
- W2952208216 hasAuthorship W2952208216A5020580195 @default.
- W2952208216 hasAuthorship W2952208216A5069169048 @default.
- W2952208216 hasAuthorship W2952208216A5087673923 @default.
- W2952208216 hasBestOaLocation W29522082161 @default.
- W2952208216 hasConcept C114614502 @default.
- W2952208216 hasConcept C117160843 @default.
- W2952208216 hasConcept C121332964 @default.
- W2952208216 hasConcept C134306372 @default.
- W2952208216 hasConcept C138885662 @default.
- W2952208216 hasConcept C154945302 @default.
- W2952208216 hasConcept C162319229 @default.
- W2952208216 hasConcept C163716315 @default.
- W2952208216 hasConcept C171686336 @default.
- W2952208216 hasConcept C173608175 @default.
- W2952208216 hasConcept C188045654 @default.
- W2952208216 hasConcept C199360897 @default.
- W2952208216 hasConcept C2778119891 @default.
- W2952208216 hasConcept C2779808786 @default.
- W2952208216 hasConcept C2781357197 @default.
- W2952208216 hasConcept C33923547 @default.
- W2952208216 hasConcept C41008148 @default.
- W2952208216 hasConcept C41895202 @default.
- W2952208216 hasConcept C500882744 @default.
- W2952208216 hasConcept C56372850 @default.
- W2952208216 hasConcept C62520636 @default.
- W2952208216 hasConcept C74193536 @default.
- W2952208216 hasConcept C80444323 @default.
- W2952208216 hasConceptScore W2952208216C114614502 @default.
- W2952208216 hasConceptScore W2952208216C117160843 @default.
- W2952208216 hasConceptScore W2952208216C121332964 @default.
- W2952208216 hasConceptScore W2952208216C134306372 @default.
- W2952208216 hasConceptScore W2952208216C138885662 @default.
- W2952208216 hasConceptScore W2952208216C154945302 @default.
- W2952208216 hasConceptScore W2952208216C162319229 @default.
- W2952208216 hasConceptScore W2952208216C163716315 @default.
- W2952208216 hasConceptScore W2952208216C171686336 @default.
- W2952208216 hasConceptScore W2952208216C173608175 @default.
- W2952208216 hasConceptScore W2952208216C188045654 @default.
- W2952208216 hasConceptScore W2952208216C199360897 @default.
- W2952208216 hasConceptScore W2952208216C2778119891 @default.
- W2952208216 hasConceptScore W2952208216C2779808786 @default.
- W2952208216 hasConceptScore W2952208216C2781357197 @default.
- W2952208216 hasConceptScore W2952208216C33923547 @default.
- W2952208216 hasConceptScore W2952208216C41008148 @default.
- W2952208216 hasConceptScore W2952208216C41895202 @default.
- W2952208216 hasConceptScore W2952208216C500882744 @default.
- W2952208216 hasConceptScore W2952208216C56372850 @default.
- W2952208216 hasConceptScore W2952208216C62520636 @default.
- W2952208216 hasConceptScore W2952208216C74193536 @default.
- W2952208216 hasConceptScore W2952208216C80444323 @default.
- W2952208216 hasLocation W29522082161 @default.
- W2952208216 hasOpenAccess W2952208216 @default.
- W2952208216 hasPrimaryLocation W29522082161 @default.
- W2952208216 hasRelatedWork W1997955449 @default.
- W2952208216 hasRelatedWork W2076165488 @default.
- W2952208216 hasRelatedWork W2133705126 @default.
- W2952208216 hasRelatedWork W2274687532 @default.
- W2952208216 hasRelatedWork W2529865518 @default.
- W2952208216 hasRelatedWork W2535419693 @default.
- W2952208216 hasRelatedWork W2952208216 @default.
- W2952208216 hasRelatedWork W3139495005 @default.
- W2952208216 hasRelatedWork W4232486587 @default.
- W2952208216 hasRelatedWork W4256483032 @default.
- W2952208216 isParatext "false" @default.
- W2952208216 isRetracted "false" @default.
- W2952208216 magId "2952208216" @default.
- W2952208216 workType "article" @default.