GoldenGolden
Advanced Search
Non-negative matrix factorization via archetypal analysis

Non-negative matrix factorization via archetypal analysis

An approach to non-negative matrix factorization that does not require data to be separable and provides a generally unique decomposition.

All edits

Edits on 19 Mar, 2019
Jude Gomila
Jude Gomila approved a suggestion from Golden's AI on 19 Mar, 2019
Edits made to:
Article (+8/-8 characters)
Article

Unlike the original archetypal analysis method developed by Cutler and Breiman, NMF via archetypal analysis does not require the data in a given data setdata set to be separable. The method aims to optimize the trade-off between two objectives:

Edits on 12 Mar, 2019
Daniel Frumkin"Wrote article"
Daniel Frumkin edited on 12 Mar, 2019
Edits made to:
Article (+1044 characters)
Table (+2 rows) (+4 cells) (+42 characters)
Article

Non-negative matrix factorization via archetypal analysis is named after two well-known techniques of statistics and unsupervised learning, non-negative matrix factorization (NMF) and archetypal analysis (AA).

Unlike the original archetypal analysis method developed by Cutler and Breiman, NMF via archetypal analysis does not require the data in a given data set to be separable. The method aims to optimize the trade-off between two objectives:

  • Minimizing the distance of the data points from the convex envelope of archetypes (which can be interpreted as an empirical risk); and
  • Minimizing the distance of the the archetypes from the convex envelope of data (which can be interpreted as a data-dependent regularization).

NMF via archetypal analysis introduces a 'uniqueness condition' on the data which is necessary for exactly recovering the archetypes from noiseless data. The approach requires solving a non-convex optimization problem, but early experiments showed that the standard optimization methods succeeded in finding good solutions.

Table

Name
Role
Related Golden topics

Andrea Montanari

Creator

Hamid Javadi

Creator

Edits on 10 Mar, 2019
Daniel Frumkin
Daniel Frumkin edited on 10 Mar, 2019
Edits made to:
Description (+33/-33 characters)
Table (+2 rows) (+8 cells) (+355 characters)
Related Topics (+2 topics)
Topic thumbnail

Non-negative matrix factorization via archetypal analysis

An approach to non-negative matrix factorizationnon-negative matrix factorization that does not require data to be separable and provides a generally unique decomposition.

Table

Title
Author
Link
Type

A geometric approach to archetypal analysis and non-negative matrix factorization

Anil Damle, Yuekai Sun

PDF

Fast and Robust Archetypal Analysis for Representation Learning

Yuansi Chen, Julien Mairal, Zaid Harchaoui

PDF

Related Topics
Edits on 9 Mar, 2019
Daniel Frumkin"Created page"
Daniel Frumkin edited on 9 Mar, 2019
Edits made to:
Description (+138 characters)
Table (+2 rows) (+8 cells) (+253 characters)
Categories (+2 topics)
Related Topics (+5 topics)
Topic thumbnail

Non-negative matrix factorization via archetypal analysis

An approach to non-negative matrix factorization that does not require data to be separable and provides a generally unique decomposition.

Table

Title
Author
Link
Type

Archetypal analysis for machine learning

Morten Morup, Lars Kai Hansen

PDF

Non-negative Matrix Factorization via Archetypal Analysis

Hamid Javadi, Andrea Montanari

PDF

Categories
Related Topics
Daniel Frumkin"Initial topic creation"
Daniel Frumkin created this topic on 9 Mar, 2019
Edits made to:
Topic thumbnail

 Non-negative matrix factorization via archetypal analysis

An approach to non-negative matrix factorization that does not require data to be separable and provides a generally unique decomposition.

Golden logo
Text is available under the Creative Commons Attribution-ShareAlike 4.0; additional terms apply. By using this site, you agree to our Terms & Conditions.