Ändra sökning
RefereraExporteraLänk till posten
Permanent länk

Direktlänk
Referera
Referensformat
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Annat format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annat språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf
Determinantal point processes for mini-batch diversification
KTH, Skolan för datavetenskap och kommunikation (CSC), Robotik, perception och lärande, RPL.ORCID-id: 0000-0002-5750-9655
2017 (Engelska)Ingår i: Uncertainty in Artificial Intelligence - Proceedings of the 33rd Conference, UAI 2017, AUAI Press Corvallis , 2017Konferensbidrag (Refereegranskat)
Abstract [en]

We study a mini-batch diversification scheme for stochastic gradient descent (SGD). While classical SGD relies on uniformly sampling data points to form a mini-batch, we propose a non-uniform sampling scheme based on the Determinantal Point Process (DPP). The DPP relies on a similarity measure between data points and gives low probabilities to mini-batches which contain redundant data, and higher probabilities to mini-batches with more diverse data. This simultaneously balances the data and leads to stochastic gradients with lower variance. We term this approach Diversified Mini-Batch SGD (DM-SGD). We show that regular SGD and a biased version of stratified sampling emerge as special cases. Furthermore, DM-SGD generalizes stratified sampling to cases where no discrete features exist to bin the data into groups. We show experimentally that our method results more interpretable and diverse features in unsupervised setups, and in better classification accuracies in supervised setups.

Ort, förlag, år, upplaga, sidor
AUAI Press Corvallis , 2017.
Nationell ämneskategori
Sannolikhetsteori och statistik
Identifikatorer
URN: urn:nbn:se:kth:diva-218565Scopus ID: 2-s2.0-85031095282OAI: oai:DiVA.org:kth-218565DiVA, id: diva2:1161321
Konferens
33rd Conference on Uncertainty in Artificial Intelligence, UAI 2017, Sydney, Australia, 11 August 2017 through 15 August 2017
Anmärkning

QC 20171129

Tillgänglig från: 2017-11-29 Skapad: 2017-11-29 Senast uppdaterad: 2017-11-29Bibliografiskt granskad

Open Access i DiVA

Fulltext saknas i DiVA

Scopus

Personposter BETA

Kjellström, Hedvig

Sök vidare i DiVA

Av författaren/redaktören
Kjellström, Hedvig
Av organisationen
Robotik, perception och lärande, RPL
Sannolikhetsteori och statistik

Sök vidare utanför DiVA

GoogleGoogle Scholar

urn-nbn

Altmetricpoäng

urn-nbn
Totalt: 8 träffar
RefereraExporteraLänk till posten
Permanent länk

Direktlänk
Referera
Referensformat
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Annat format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annat språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf