https://github.com/lmjohns3/kohonen

Kohonen vector quantizers (SOM, NG, GNG)
https://github.com/lmjohns3/kohonen

Last synced: about 2 months ago
JSON representation

Kohonen vector quantizers (SOM, NG, GNG)

• Host: GitHub
• URL: https://github.com/lmjohns3/kohonen
• Owner: lmjohns3
• License: mit
• Created: 2011-05-20T22:15:17.000Z (about 13 years ago)
• Default Branch: master
• Last Pushed: 2018-06-04T04:35:58.000Z (about 6 years ago)
• Last Synced: 2024-03-21T00:08:43.905Z (3 months ago)
• Language: Python
• Homepage: http://kohonen.rtfd.org
• Size: 105 KB
• Stars: 72
• Watchers: 3
• Forks: 15
• Open Issues: 2
• Metadata Files:
  • Readme: README.rst
  • License: LICENSE

Lists

• awesome-TDA - Kohonen - Kohonen-style vector quantizers: Self-Organizing Map (SOM), Neural Gas, and Growing Neural Gas. (Frameworks and Libs / Python)
• alex-mikhalev-awesome-stars - kohonen - Kohonen vector quantizers (SOM, NG, GNG) (Python)

README

        
kohonen

=======

This module contains some basic implementations of Kohonen-style vector

quantizers: Self-Organizing Map (SOM), Neural Gas, and Growing Neural Gas.

Kohonen-style vector quantizers use some sort of explicitly specified topology

to encourage good separation among prototype "neurons".

Vector quantizers are useful for learning discrete representations of a

distribution over continuous space, based solely on samples drawn from the

distribution. This process is also generally known as density estimation.

The source distribution includes an interactive test module that uses PyGTK and

Cairo to render a set of quantizers that move around in real time as samples are

drawn from a known distribution and fed to the quantizers. Run this test with::

    python kohonen_test.py

Documentation (currently a bit sparse) lives at http://pythonhosted.org/kohonen.

Have fun!