Discrete Sparse Coding

Sparse coding algorithms with continuous latent variables have been the subject of a large number of studies. However, discrete latent spaces for sparse coding have been largely ignored. In this work, we study sparse coding with latents described by discrete instead of continuous prior distributions...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Neural computation Ročník 29; číslo 11; s. 2979
Hlavní autori: Exarchakis, Georgios, Lücke, Jörg
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: United States 01.11.2017
ISSN:1530-888X, 1530-888X
On-line prístup:Zistit podrobnosti o prístupe
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Abstract Sparse coding algorithms with continuous latent variables have been the subject of a large number of studies. However, discrete latent spaces for sparse coding have been largely ignored. In this work, we study sparse coding with latents described by discrete instead of continuous prior distributions. We consider the general case in which the latents (while being sparse) can take on any value of a finite set of possible values and in which we learn the prior probability of any value from data. This approach can be applied to any data generated by discrete causes, and it can be applied as an approximation of continuous causes. As the prior probabilities are learned, the approach then allows for estimating the prior shape without assuming specific functional forms. To efficiently train the parameters of our probabilistic generative model, we apply a truncated expectation-maximization approach (expectation truncation) that we modify to work with a general discrete prior. We evaluate the performance of the algorithm by applying it to a variety of tasks: (1) we use artificial data to verify that the algorithm can recover the generating parameters from a random initialization, (2) use image patches of natural images and discuss the role of the prior for the extraction of image components, (3) use extracellular recordings of neurons to present a novel method of analysis for spiking neurons that includes an intuitive discretization strategy, and (4) apply the algorithm on the task of encoding audio waveforms of human speech. The diverse set of numerical experiments presented in this letter suggests that discrete sparse coding algorithms can scale efficiently to work with realistic data sets and provide novel statistical quantities to describe the structure of the data.
AbstractList Sparse coding algorithms with continuous latent variables have been the subject of a large number of studies. However, discrete latent spaces for sparse coding have been largely ignored. In this work, we study sparse coding with latents described by discrete instead of continuous prior distributions. We consider the general case in which the latents (while being sparse) can take on any value of a finite set of possible values and in which we learn the prior probability of any value from data. This approach can be applied to any data generated by discrete causes, and it can be applied as an approximation of continuous causes. As the prior probabilities are learned, the approach then allows for estimating the prior shape without assuming specific functional forms. To efficiently train the parameters of our probabilistic generative model, we apply a truncated expectation-maximization approach (expectation truncation) that we modify to work with a general discrete prior. We evaluate the performance of the algorithm by applying it to a variety of tasks: (1) we use artificial data to verify that the algorithm can recover the generating parameters from a random initialization, (2) use image patches of natural images and discuss the role of the prior for the extraction of image components, (3) use extracellular recordings of neurons to present a novel method of analysis for spiking neurons that includes an intuitive discretization strategy, and (4) apply the algorithm on the task of encoding audio waveforms of human speech. The diverse set of numerical experiments presented in this letter suggests that discrete sparse coding algorithms can scale efficiently to work with realistic data sets and provide novel statistical quantities to describe the structure of the data.Sparse coding algorithms with continuous latent variables have been the subject of a large number of studies. However, discrete latent spaces for sparse coding have been largely ignored. In this work, we study sparse coding with latents described by discrete instead of continuous prior distributions. We consider the general case in which the latents (while being sparse) can take on any value of a finite set of possible values and in which we learn the prior probability of any value from data. This approach can be applied to any data generated by discrete causes, and it can be applied as an approximation of continuous causes. As the prior probabilities are learned, the approach then allows for estimating the prior shape without assuming specific functional forms. To efficiently train the parameters of our probabilistic generative model, we apply a truncated expectation-maximization approach (expectation truncation) that we modify to work with a general discrete prior. We evaluate the performance of the algorithm by applying it to a variety of tasks: (1) we use artificial data to verify that the algorithm can recover the generating parameters from a random initialization, (2) use image patches of natural images and discuss the role of the prior for the extraction of image components, (3) use extracellular recordings of neurons to present a novel method of analysis for spiking neurons that includes an intuitive discretization strategy, and (4) apply the algorithm on the task of encoding audio waveforms of human speech. The diverse set of numerical experiments presented in this letter suggests that discrete sparse coding algorithms can scale efficiently to work with realistic data sets and provide novel statistical quantities to describe the structure of the data.
Sparse coding algorithms with continuous latent variables have been the subject of a large number of studies. However, discrete latent spaces for sparse coding have been largely ignored. In this work, we study sparse coding with latents described by discrete instead of continuous prior distributions. We consider the general case in which the latents (while being sparse) can take on any value of a finite set of possible values and in which we learn the prior probability of any value from data. This approach can be applied to any data generated by discrete causes, and it can be applied as an approximation of continuous causes. As the prior probabilities are learned, the approach then allows for estimating the prior shape without assuming specific functional forms. To efficiently train the parameters of our probabilistic generative model, we apply a truncated expectation-maximization approach (expectation truncation) that we modify to work with a general discrete prior. We evaluate the performance of the algorithm by applying it to a variety of tasks: (1) we use artificial data to verify that the algorithm can recover the generating parameters from a random initialization, (2) use image patches of natural images and discuss the role of the prior for the extraction of image components, (3) use extracellular recordings of neurons to present a novel method of analysis for spiking neurons that includes an intuitive discretization strategy, and (4) apply the algorithm on the task of encoding audio waveforms of human speech. The diverse set of numerical experiments presented in this letter suggests that discrete sparse coding algorithms can scale efficiently to work with realistic data sets and provide novel statistical quantities to describe the structure of the data.
Author Lücke, Jörg
Exarchakis, Georgios
Author_xml – sequence: 1
  givenname: Georgios
  surname: Exarchakis
  fullname: Exarchakis, Georgios
  email: georgios.exarchakis@uol.de
  organization: Machine Learning Lab, Cluster of Excellence Hearing4all and Department for Medical Physics and Acoustics, Carl-von-Ossietzky University Oldenburg, 26111 Oldenburg, Germany georgios.exarchakis@uol.de
– sequence: 2
  givenname: Jörg
  surname: Lücke
  fullname: Lücke, Jörg
  email: joerg.luecke@uol.de
  organization: Machine Learning Lab, Cluster of Excellence Hearing4all and Department for Medical Physics and Acoustics, Carl-von-Ossietzky University Oldenburg, 26111 Oldenburg, Germany joerg.luecke@uol.de
BackLink https://www.ncbi.nlm.nih.gov/pubmed/28957027$$D View this record in MEDLINE/PubMed
BookMark eNpNjztLxEAUhQdZcR_aiaVsaRO9d543pcQnLFioYDeMkxuJbB5mksJ_74IrWJ1TfHycsxSztmtZiFOES0Qrr1qOnQ8eENAciAUaBRkRvc3-9blYpvQJABbBHIm5pNw4kG4hzm7qFAceef3chyHxuujKuv04FodV2CY-2edKvN7dvhQP2ebp_rG43mRRaxozDsExlcpZqShnayrDgUIOUZN1rAjYBWPBqWgrnSOBthGcIQJpoAxyJS5-vf3QfU2cRt_s9vB2G1rupuQx10ZKhYg79HyPTu8Nl74f6iYM3_7vi_wBUYRKSA
CitedBy_id crossref_primary_10_3390_math11122674
crossref_primary_10_1088_2634_4386_ac970d
ContentType Journal Article
DBID NPM
7X8
DOI 10.1162/neco_a_01015
DatabaseName PubMed
MEDLINE - Academic
DatabaseTitle PubMed
MEDLINE - Academic
DatabaseTitleList MEDLINE - Academic
PubMed
Database_xml – sequence: 1
  dbid: NPM
  name: PubMed
  url: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed
  sourceTypes: Index Database
– sequence: 2
  dbid: 7X8
  name: MEDLINE - Academic
  url: https://search.proquest.com/medline
  sourceTypes: Aggregation Database
DeliveryMethod no_fulltext_linktorsrc
Discipline Computer Science
EISSN 1530-888X
ExternalDocumentID 28957027
Genre Research Support, Non-U.S. Gov't
Journal Article
GroupedDBID ---
-~X
.4S
.DC
0R~
123
36B
4.4
41~
53G
6IK
AAJGR
AALMD
AAYOK
ABAZT
ABDBF
ABDNZ
ABEFU
ABIVO
ABJNI
ACGFO
ACUHS
ACYGS
ADIYS
ADMLS
AEGXH
AEILP
AENEX
AIAGR
ALMA_UNASSIGNED_HOLDINGS
ARCSS
AVWKF
AZFZN
BEFXN
BFFAM
BGNUA
BKEBE
BPEOZ
CAG
COF
CS3
DU5
EAP
EAS
EBC
EBD
EBS
ECS
EDO
EJD
EMB
EMK
EMOBN
EPL
EPS
EST
ESX
F5P
FEDTE
FNEHJ
HVGLF
HZ~
H~9
I-F
IPLJI
JAVBF
MCG
MINIK
MKJ
NPM
O9-
OCL
P2P
PK0
PQQKQ
RMI
SV3
TUS
WG8
WH7
XJE
ZWS
7X8
ABUFD
ABVLG
AMVHM
ID FETCH-LOGICAL-c448t-eaa7e8d3762389e65f5ea8a90c4867e380e7a56073c6f4918046c075880250da2
IEDL.DBID 7X8
ISICitedReferencesCount 8
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000413292600005&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 1530-888X
IngestDate Sun Nov 09 13:59:25 EST 2025
Thu Apr 03 07:05:38 EDT 2025
IsDoiOpenAccess false
IsOpenAccess true
IsPeerReviewed true
IsScholarly true
Issue 11
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-c448t-eaa7e8d3762389e65f5ea8a90c4867e380e7a56073c6f4918046c075880250da2
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
OpenAccessLink https://direct.mit.edu/neco/article-pdf/29/11/2979/1026079/neco_a_01015.pdf
PMID 28957027
PQID 1945223111
PQPubID 23479
ParticipantIDs proquest_miscellaneous_1945223111
pubmed_primary_28957027
PublicationCentury 2000
PublicationDate 2017-11-00
20171101
PublicationDateYYYYMMDD 2017-11-01
PublicationDate_xml – month: 11
  year: 2017
  text: 2017-11-00
PublicationDecade 2010
PublicationPlace United States
PublicationPlace_xml – name: United States
PublicationTitle Neural computation
PublicationTitleAlternate Neural Comput
PublicationYear 2017
SSID ssj0006105
Score 2.299804
Snippet Sparse coding algorithms with continuous latent variables have been the subject of a large number of studies. However, discrete latent spaces for sparse coding...
SourceID proquest
pubmed
SourceType Aggregation Database
Index Database
StartPage 2979
Title Discrete Sparse Coding
URI https://www.ncbi.nlm.nih.gov/pubmed/28957027
https://www.proquest.com/docview/1945223111
Volume 29
WOSCitedRecordID wos000413292600005&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText
inHoldings 1
isFullTextHit
isPrint
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwpV1LS8NAEB7UevBifVXriwhelzbZ7CMnkWrxoKXgg9zCupmFXpJoq7_f2TzoSRC85BbIJjvzfd_O5BuA6yjCRHOXMynimMXGGEYgGDG0zuXev0zVFd23RzWb6TRN5u2B27Jtq-xyYp2o89L6M_IRiW2iCpxC86b6YH5qlK-utiM0NqHHicr4li6Vrt3CZdPCSEE9ZqT00q7xXUajgsRdZjLvsCZ-J5c1yEz7_328Pdht6WVw2-yHfdjA4gD63eiGoI3kQxjcLShfEGEOniuSthhMSg9iR_A6vX-ZPLB2RAKzpKtWDI1RqHPKEgS9CUrhBBptkrH1TnrI9RiVIVKjuJUuTkJNctgSS6CoJe6Tm2gAW0VZ4AkEEkn6hNG7jYWLQxKXjgsrtbJcUj7PwyFcdSvPaAv6uoIpsPxaZuu1D-G4eX1Z1XhlZKTnhCLpe_qHu89gJ_KgWf_pdw49RwGIF7Btv1eL5edl_W3pOps__QBT4Kuq
linkProvider ProQuest
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Discrete+Sparse+Coding&rft.jtitle=Neural+computation&rft.au=Exarchakis%2C+Georgios&rft.au=L%C3%BCcke%2C+J%C3%B6rg&rft.date=2017-11-01&rft.issn=1530-888X&rft.eissn=1530-888X&rft.volume=29&rft.issue=11&rft.spage=2979&rft_id=info:doi/10.1162%2Fneco_a_01015&rft.externalDBID=NO_FULL_TEXT
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1530-888X&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1530-888X&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1530-888X&client=summon