Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Determinantal point processes for mini-batch diversification
KTH, School of Computer Science and Communication (CSC), Robotics, perception and learning, RPL.ORCID iD: 0000-0002-5750-9655
2017 (English)In: Uncertainty in Artificial Intelligence - Proceedings of the 33rd Conference, UAI 2017, AUAI Press Corvallis , 2017Conference paper (Refereed)
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.

Place, publisher, year, edition, pages
AUAI Press Corvallis , 2017.
National Category
Probability Theory and Statistics
Identifiers
URN: urn:nbn:se:kth:diva-218565Scopus ID: 2-s2.0-85031095282OAI: oai:DiVA.org:kth-218565DiVA, id: diva2:1161321
Conference
33rd Conference on Uncertainty in Artificial Intelligence, UAI 2017, Sydney, Australia, 11 August 2017 through 15 August 2017
Note

QC 20171129

Available from: 2017-11-29 Created: 2017-11-29 Last updated: 2017-11-29Bibliographically approved

Open Access in DiVA

No full text in DiVA

Scopus

Authority records BETA

Kjellström, Hedvig

Search in DiVA

By author/editor
Kjellström, Hedvig
By organisation
Robotics, perception and learning, RPL
Probability Theory and Statistics

Search outside of DiVA

GoogleGoogle Scholar

urn-nbn

Altmetric score

urn-nbn
Total: 8 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf