Here's how NATURE.COM makes money* and how much!

*Please read our disclaimer before using our estimates.
Loading...

NATURE . COM {}

  1. Analyzed Page
  2. Matching Content Categories
  3. CMS
  4. Monthly Traffic Estimate
  5. How Does Nature.com Make Money
  6. How Much Does Nature.com Make
  7. Keywords
  8. Topics
  9. Questions
  10. Schema
  11. Social Networks
  12. External Links
  13. Analytics And Tracking
  14. Libraries
  15. Hosting Providers
  16. CDN Services

We are analyzing https://www.nature.com/articles/44565.

Title:
Learning the parts of objects by non-negative matrix factorization | Nature
Description:
Is perception of the whole based on perception of its parts? There is psychological1 and physiological2,3 evidence for parts-based representations in the brain, and certain computational theories of object recognition rely on such representations4,5. But little is known about how brains or computers might learn the parts of objects. Here we demonstrate an algorithm for non-negative matrix factorization that is able to learn parts of faces and semantic features of text. This is in contrast to other methods, such as principal components analysis and vector quantization, that learn holistic, not parts-based, representations. Non-negative matrix factorization is distinguished from the other methods by its use of non-negativity constraints. These constraints lead to a parts-based representation because they allow only additive, not subtractive, combinations. When non-negative matrix factorization is implemented as a neural network, parts-based representations emerge by virtue of two properties: the firing rates of neurons are never negative and synaptic strengths do not change sign.
Website Age:
30 years and 10 months (reg. 1994-08-11).

Matching Content Categories {πŸ“š}

  • Education
  • Technology & Computing
  • Mobile Technology & AI

Content Management System {πŸ“}

What CMS is nature.com built with?

Custom-built

No common CMS systems were detected on Nature.com, and no known web development framework was identified.

Traffic Estimate {πŸ“ˆ}

What is the average monthly size of nature.com audience?

πŸŒ† Monumental Traffic: 20M - 50M visitors per month


Based on our best estimate, this website will receive around 41,362,249 visitors per month in the current month.

check SE Ranking
check Ahrefs
check Similarweb
check Ubersuggest
check Semrush

How Does Nature.com Make Money? {πŸ’Έ}


Display Ads {🎯}


The website utilizes display ads within its content to generate revenue. Check the next section for further revenue estimates.

Ads are managed by yourbow.com. Particular relationships are as follows:

Direct Advertisers (10)
google.com, pmc.com, doceree.com, yourbow.com, audienciad.com, onlinemediasolutions.com, advibe.media, aps.amazon.com, getmediamx.com, onomagic.com

Reseller Advertisers (38)
conversantmedia.com, rubiconproject.com, pubmatic.com, appnexus.com, openx.com, smartadserver.com, lijit.com, sharethrough.com, video.unrulymedia.com, google.com, yahoo.com, triplelift.com, onetag.com, sonobi.com, contextweb.com, 33across.com, indexexchange.com, media.net, themediagrid.com, adform.com, richaudience.com, sovrn.com, improvedigital.com, freewheel.tv, smaato.com, yieldmo.com, amxrtb.com, adyoulike.com, adpone.com, criteo.com, smilewanted.com, 152media.info, e-planning.net, smartyads.com, loopme.com, opera.com, mediafuse.com, betweendigital.com

How Much Does Nature.com Make? {πŸ’°}


Display Ads {🎯}

$521,200 per month
Our estimates place Nature.com's monthly online earnings from display ads at $347,485 to $955,583.

Keywords {πŸ”}

article, google, scholar, nature, access, neural, content, cas, pubmed, parts, nonnegative, representations, cookies, analysis, data, matrix, factorization, recognition, ads, privacy, information, learn, learning, objects, seung, based, partsbased, object, algorithm, open, press, math, scientific, advertising, subscribe, lee, sebastian, brain, methods, institution, online, buy, language, networks, psychol, cortex, visual, rev, theory, cambridge,

Topics {βœ’οΈ}

nature portfolio permissions reprints privacy policy author information authors individual-specific template-based representations advertising natural language processing social media nature 381 nature 401 nature grip-load force couplings bayesian-based iterative method parts-based representations emerge springerlink instant access permissions personal data negative factor analysis data protection incomplete data unsupervised neural networks principal components analysis negative matrix factorization neural network parts-based representations privacy independent component representations infant crawling based find parts bartlett adaptive algorithm based object recognition rely european economic area institutional subscriptions read single units high-level vision grolier encyclopedia corpus abnormal interlimb coordination motor developmental delay hybrid-sindy michael accepting optional cookies visual object recognition perceiving visual surfaces explore content subscription content journals search log form cognitive maps human image understanding em-type algorithms parts-based representation οΏ½wake-sleep” algorithm

Questions {❓}

  • Is perception of the whole based on perception of its parts?
  • What is the goal of sensory coding?

Schema {πŸ—ΊοΈ}

WebPage:
      mainEntity:
         headline:Learning the parts of objects by non-negative matrix factorization
         description:Is perception of the whole based on perception of its parts? There is psychological1 and physiological2,3 evidence for parts-based representations in the brain, and certain computational theories of object recognition rely on such representations4,5. But little is known about how brains or computers might learn the parts of objects. Here we demonstrate an algorithm for non-negative matrix factorization that is able to learn parts of faces and semantic features of text. This is in contrast to other methods, such as principal components analysis and vector quantization, that learn holistic, not parts-based, representations. Non-negative matrix factorization is distinguished from the other methods by its use of non-negativity constraints. These constraints lead to a parts-based representation because they allow only additive, not subtractive, combinations. When non-negative matrix factorization is implemented as a neural network, parts-based representations emerge by virtue of two properties: the firing rates of neurons are never negative and synaptic strengths do not change sign.
         datePublished:
         dateModified:
         pageStart:788
         pageEnd:791
         sameAs:https://doi.org/10.1038/44565
         keywords:
            Science
            Humanities and Social Sciences
            multidisciplinary
         image:
            https://media.springernature.com/lw1200/springer-static/image/art%3A10.1038%2F44565/MediaObjects/41586_1999_Article_BF44565_Fig1_HTML.gif
            https://media.springernature.com/lw1200/springer-static/image/art%3A10.1038%2F44565/MediaObjects/41586_1999_Article_BF44565_Fig2_HTML.gif
            https://media.springernature.com/lw1200/springer-static/image/art%3A10.1038%2F44565/MediaObjects/41586_1999_Article_BF44565_Fig3_HTML.gif
            https://media.springernature.com/lw1200/springer-static/image/art%3A10.1038%2F44565/MediaObjects/41586_1999_Article_BF44565_Fig4_HTML.gif
         isPartOf:
            name:Nature
            issn:
               1476-4687
               0028-0836
            volumeNumber:401
            type:
               Periodical
               PublicationVolume
         publisher:
            name:Nature Publishing Group UK
            logo:
               url:https://www.springernature.com/app-sn/public/images/logo-springernature.png
               type:ImageObject
            type:Organization
         author:
               name:Daniel D. Lee
               affiliation:
                     name:Bell Laboratories, Lucent Technologies
                     address:
                        name:Bell Laboratories, Lucent Technologies , Murray Hill, USA
                        type:PostalAddress
                     type:Organization
               type:Person
               name:H. Sebastian Seung
               affiliation:
                     name:Bell Laboratories, Lucent Technologies
                     address:
                        name:Bell Laboratories, Lucent Technologies , Murray Hill, USA
                        type:PostalAddress
                     type:Organization
                     name:Massachusetts Institute of Technology
                     address:
                        name:Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology, Cambridge, USA
                        type:PostalAddress
                     type:Organization
               type:Person
         isAccessibleForFree:
         hasPart:
            isAccessibleForFree:
            cssSelector:.main-content
            type:WebPageElement
         type:ScholarlyArticle
      context:https://schema.org
ScholarlyArticle:
      headline:Learning the parts of objects by non-negative matrix factorization
      description:Is perception of the whole based on perception of its parts? There is psychological1 and physiological2,3 evidence for parts-based representations in the brain, and certain computational theories of object recognition rely on such representations4,5. But little is known about how brains or computers might learn the parts of objects. Here we demonstrate an algorithm for non-negative matrix factorization that is able to learn parts of faces and semantic features of text. This is in contrast to other methods, such as principal components analysis and vector quantization, that learn holistic, not parts-based, representations. Non-negative matrix factorization is distinguished from the other methods by its use of non-negativity constraints. These constraints lead to a parts-based representation because they allow only additive, not subtractive, combinations. When non-negative matrix factorization is implemented as a neural network, parts-based representations emerge by virtue of two properties: the firing rates of neurons are never negative and synaptic strengths do not change sign.
      datePublished:
      dateModified:
      pageStart:788
      pageEnd:791
      sameAs:https://doi.org/10.1038/44565
      keywords:
         Science
         Humanities and Social Sciences
         multidisciplinary
      image:
         https://media.springernature.com/lw1200/springer-static/image/art%3A10.1038%2F44565/MediaObjects/41586_1999_Article_BF44565_Fig1_HTML.gif
         https://media.springernature.com/lw1200/springer-static/image/art%3A10.1038%2F44565/MediaObjects/41586_1999_Article_BF44565_Fig2_HTML.gif
         https://media.springernature.com/lw1200/springer-static/image/art%3A10.1038%2F44565/MediaObjects/41586_1999_Article_BF44565_Fig3_HTML.gif
         https://media.springernature.com/lw1200/springer-static/image/art%3A10.1038%2F44565/MediaObjects/41586_1999_Article_BF44565_Fig4_HTML.gif
      isPartOf:
         name:Nature
         issn:
            1476-4687
            0028-0836
         volumeNumber:401
         type:
            Periodical
            PublicationVolume
      publisher:
         name:Nature Publishing Group UK
         logo:
            url:https://www.springernature.com/app-sn/public/images/logo-springernature.png
            type:ImageObject
         type:Organization
      author:
            name:Daniel D. Lee
            affiliation:
                  name:Bell Laboratories, Lucent Technologies
                  address:
                     name:Bell Laboratories, Lucent Technologies , Murray Hill, USA
                     type:PostalAddress
                  type:Organization
            type:Person
            name:H. Sebastian Seung
            affiliation:
                  name:Bell Laboratories, Lucent Technologies
                  address:
                     name:Bell Laboratories, Lucent Technologies , Murray Hill, USA
                     type:PostalAddress
                  type:Organization
                  name:Massachusetts Institute of Technology
                  address:
                     name:Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology, Cambridge, USA
                     type:PostalAddress
                  type:Organization
            type:Person
      isAccessibleForFree:
      hasPart:
         isAccessibleForFree:
         cssSelector:.main-content
         type:WebPageElement
["Periodical","PublicationVolume"]:
      name:Nature
      issn:
         1476-4687
         0028-0836
      volumeNumber:401
Organization:
      name:Nature Publishing Group UK
      logo:
         url:https://www.springernature.com/app-sn/public/images/logo-springernature.png
         type:ImageObject
      name:Bell Laboratories, Lucent Technologies
      address:
         name:Bell Laboratories, Lucent Technologies , Murray Hill, USA
         type:PostalAddress
      name:Bell Laboratories, Lucent Technologies
      address:
         name:Bell Laboratories, Lucent Technologies , Murray Hill, USA
         type:PostalAddress
      name:Massachusetts Institute of Technology
      address:
         name:Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology, Cambridge, USA
         type:PostalAddress
ImageObject:
      url:https://www.springernature.com/app-sn/public/images/logo-springernature.png
Person:
      name:Daniel D. Lee
      affiliation:
            name:Bell Laboratories, Lucent Technologies
            address:
               name:Bell Laboratories, Lucent Technologies , Murray Hill, USA
               type:PostalAddress
            type:Organization
      name:H. Sebastian Seung
      affiliation:
            name:Bell Laboratories, Lucent Technologies
            address:
               name:Bell Laboratories, Lucent Technologies , Murray Hill, USA
               type:PostalAddress
            type:Organization
            name:Massachusetts Institute of Technology
            address:
               name:Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology, Cambridge, USA
               type:PostalAddress
            type:Organization
PostalAddress:
      name:Bell Laboratories, Lucent Technologies , Murray Hill, USA
      name:Bell Laboratories, Lucent Technologies , Murray Hill, USA
      name:Department of Brain and Cognitive Sciences, Massachusetts Institute of Technology, Cambridge, USA
WebPageElement:
      isAccessibleForFree:
      cssSelector:.main-content

External Links {πŸ”—}(77)

Analytics and Tracking {πŸ“Š}

  • Google Tag Manager

Libraries {πŸ“š}

  • Prism.js
  • Zoom.js

Emails and Hosting {βœ‰οΈ}

Mail Servers:

  • mxa-002c5801.gslb.pphosted.com
  • mxb-002c5801.gslb.pphosted.com

Name Servers:

  • pdns1.ultradns.net
  • pdns2.ultradns.net
  • pdns3.ultradns.org
  • pdns4.ultradns.org
  • pdns5.ultradns.info
  • pdns6.ultradns.co.uk

CDN Services {πŸ“¦}

  • Crossref

4.16s.