File Download
  Links for fulltext
     (May Require Subscription)
Supplementary

Article: Tree-based machine learning performed in-memory with memristive analog CAM

TitleTree-based machine learning performed in-memory with memristive analog CAM
Authors
Issue Date2021
PublisherNature Research: Fully open access journals. The Journal's web site is located at http://www.nature.com/ncomms/index.html
Citation
Nature Communications, 2021, v. 12 n. 1, p. article no. 5806 How to Cite?
AbstractTree-based machine learning techniques, such as Decision Trees and Random Forests, are top performers in several domains as they do well with limited training datasets and offer improved interpretability compared to Deep Neural Networks (DNN). However, these models are difficult to optimize for fast inference at scale without accuracy loss in von Neumann architectures due to non-uniform memory access patterns. Recently, we proposed a novel analog content addressable memory (CAM) based on emerging memristor devices for fast look-up table operations. Here, we propose for the first time to use the analog CAM as an in-memory computational primitive to accelerate tree-based model inference. We demonstrate an efficient mapping algorithm leveraging the new analog CAM capabilities such that each root to leaf path of a Decision Tree is programmed into a row. This new in-memory compute concept for enables few-cycle model inference, dramatically increasing 103 × the throughput over conventional approaches.
Persistent Identifierhttp://hdl.handle.net/10722/305794
ISSN
2023 Impact Factor: 14.7
2023 SCImago Journal Rankings: 4.887
PubMed Central ID
ISI Accession Number ID

 

DC FieldValueLanguage
dc.contributor.authorPedretti, G-
dc.contributor.authorGraves, GE-
dc.contributor.authorSerebryakov, S-
dc.contributor.authorMAO, R-
dc.contributor.authorSheng, X-
dc.contributor.authorFoltin, M-
dc.contributor.authorLi, C-
dc.contributor.authorStrachan, JP-
dc.date.accessioned2021-10-20T10:14:25Z-
dc.date.available2021-10-20T10:14:25Z-
dc.date.issued2021-
dc.identifier.citationNature Communications, 2021, v. 12 n. 1, p. article no. 5806-
dc.identifier.issn2041-1723-
dc.identifier.urihttp://hdl.handle.net/10722/305794-
dc.description.abstractTree-based machine learning techniques, such as Decision Trees and Random Forests, are top performers in several domains as they do well with limited training datasets and offer improved interpretability compared to Deep Neural Networks (DNN). However, these models are difficult to optimize for fast inference at scale without accuracy loss in von Neumann architectures due to non-uniform memory access patterns. Recently, we proposed a novel analog content addressable memory (CAM) based on emerging memristor devices for fast look-up table operations. Here, we propose for the first time to use the analog CAM as an in-memory computational primitive to accelerate tree-based model inference. We demonstrate an efficient mapping algorithm leveraging the new analog CAM capabilities such that each root to leaf path of a Decision Tree is programmed into a row. This new in-memory compute concept for enables few-cycle model inference, dramatically increasing 103 × the throughput over conventional approaches.-
dc.languageeng-
dc.publisherNature Research: Fully open access journals. The Journal's web site is located at http://www.nature.com/ncomms/index.html-
dc.relation.ispartofNature Communications-
dc.rightsNature Communications. Copyright © Nature Research: Fully open access journals.-
dc.rightsThis work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.-
dc.titleTree-based machine learning performed in-memory with memristive analog CAM-
dc.typeArticle-
dc.identifier.emailLi, C: canl@hku.hk-
dc.identifier.authorityLi, C=rp02706-
dc.description.naturepublished_or_final_version-
dc.identifier.doi10.1038/s41467-021-25873-0-
dc.identifier.pmid34608133-
dc.identifier.pmcidPMC8490381-
dc.identifier.scopuseid_2-s2.0-85116353650-
dc.identifier.hkuros327481-
dc.identifier.volume12-
dc.identifier.issue1-
dc.identifier.spagearticle no. 5806-
dc.identifier.epagearticle no. 5806-
dc.identifier.isiWOS:000703617100028-
dc.publisher.placeUnited Kingdom-

Export via OAI-PMH Interface in XML Formats


OR


Export to Other Non-XML Formats