English

Global Collaboration through Local Interaction in Competitive Learning

Machine Learning 2019-02-12 v1 Neural and Evolutionary Computing Machine Learning

Abstract

Feature maps, that preserve the global topology of arbitrary datasets, can be formed by self-organizing competing agents. So far, it has been presumed that global interaction of agents is necessary for this process. We establish that this is not the case, and that global topology can be uncovered through strictly local interactions. Enforcing uniformity of map quality across all agents, results in an algorithm that is able to consistently uncover the global topology of diversely challenging datasets.The applicability and scalability of this approach is further tested on a large point cloud dataset, revealing a linear relation between map training time and size. The presented work not only reduces algorithmic complexity but also constitutes first step towards a distributed self organizing map.

Keywords

Cite

@article{arxiv.1902.03856,
  title  = {Global Collaboration through Local Interaction in Competitive Learning},
  author = {Abbas Siddiqui and Dionysios Georgiadis},
  journal= {arXiv preprint arXiv:1902.03856},
  year   = {2019}
}

Comments

The behavior via simulation can be viewed at: https://www.youtube.com/watch?v=lTxlVHXGO2Q

R2 v1 2026-06-23T07:37:32.889Z