NN2SQL: Let SQL Think for Neural Networks

Maximilian E. Schüle, Alfons Kemper, Thomas Neumann

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

2 Scopus citations

Abstract

Although database systems perform well in data access and manipulation, their relational model hinders data scientists from formulating machine learning algorithms in SQL. Nevertheless, we argue that modern database systems perform well for machine learning algorithms expressed in relational algebra. To overcome the barrier of the relational model, this paper shows how to transform data into a relational representation for training neural networks in SQL: We first describe building blocks for data transformation in SQL. Then, we compare an implementation for model training using array data types to the one using a relational representation in SQL-92 only. The evaluation proves the suitability of modern database systems for matrix algebra, although specialised array data types perform better than matrices in relational representation.

Original languageEnglish
Title of host publicationDatenbanksysteme fur Business, Technologie und Web, BTW 2023
EditorsBirgitta Konig-Ries, Stefanie Scherzinger, Wolfgang Lehner, Gottfried Vossen
PublisherGesellschaft fur Informatik (GI)
Pages183-194
Number of pages12
ISBN (Electronic)9783885797258
DOIs
StatePublished - 2023
Event2023 Datenbanksysteme fur Business, Technologie und Web, BTW 2023 - 2023 Database Systems for Business, Technology and Web, BTW 2023 - Dresden, Germany
Duration: 6 Mar 202310 Mar 2023

Publication series

NameLecture Notes in Informatics (LNI), Proceedings - Series of the Gesellschaft fur Informatik (GI)
VolumeP-331
ISSN (Print)1617-5468

Conference

Conference2023 Datenbanksysteme fur Business, Technologie und Web, BTW 2023 - 2023 Database Systems for Business, Technology and Web, BTW 2023
Country/TerritoryGermany
CityDresden
Period6/03/2310/03/23

Keywords

  • Automatic Differentiation
  • Neural Networks
  • SQL-92

Fingerprint

Dive into the research topics of 'NN2SQL: Let SQL Think for Neural Networks'. Together they form a unique fingerprint.

Cite this