Recurrent Self-Organizing Maps

This is package contains useful function to simulate and train Recurrent Self-Organizing Maps (RSOM) under MATLAB computational platform. It currently depends on Neural Networks Toolbox of MATLAB, but I am planning to remove these dependencies.Author of this package is Amir Reza Saffari Azar Alamdari. This toolbox is created and developed by personal interests of the author on unsupervised learning models for time series and is provided to other users as an open source free software under GNU GPL. Feel free to copy, modify, and distribute it according to the license. To obtain the latest version of GPL, please refer to Note that MATLAB itself is a commercial software provided by The MathWorks, Inc.

Download the latest version from here:

Version 1.0: (Release date: July 16, 2005) RSOM Toolbox

7 thoughts on “Recurrent Self-Organizing Maps

  1. Hi!
    I am an Engineering student in University of Coimbra, Portugal and I’m researching on some clustering methods to detect (and perhaps) improve an industrial process that depends on tens of variables collected every hour.
    I achieved some interesting results by using the SOM Toolbox for Matlab, developed by Esa Alhoniemi, Johan Himberg, Jukka Parviainen and Juha Vesanto. Now I’m trying to improve my unsupervised clustering methods by finding some time dependent relations between measures.
    I’m using your RSOM toolbox to generate neural networks with defined number of inputs and SOM size, but I have some difficulties on getting conclusions:
    – What is the best choice for my purpose: a Recursive SOM or a SOM-Structured Data model?
    – Is it possible to apply clustering algorithms directly on a trained SOM Net?
    – This kind of neural networks is stable for many inputs? (I notated that you used only one input on the examples)

    Thank you for your attention.
    Best regards,

    Edgar Ferreira, Portugal

  2. Hi. I’m about recurent SOM. I am an Engineering student in University of RUS. please, can you give me some demonstration about work RecSOM.


  3. Hi, I am researching on some time series prediction techniques on binary datasets (collected from sensors). Can you give some explaination on how the output of RSOM is demonstrated in your work?

  4. Hello,

    I downloaded RSOM from this site. I hope it is the latest version.

    I would like to use RSOM for time series clustering (self-organizing clusters). I would appreciate your detailed response and help in this. Thanks!

  5. salam agha amir man fek kardam shoma ham bayad irani bashidmishe mno dar morede karborde shabake asabi too pardzeo sot va speech rahnamayi konid?
    man reshte mohandesi seda mikhoona
    movaffagh ashid
    Nahid Saadati

  6. salam
    khaste nabashid
    manam irani hastam
    man darbary karbord U_matrix dar SOM komak mikhastam
    shoma mitonid tozih bedid U_matrix chi has va che karbordi dare?

Leave a Reply