Tackling learning intractability through topological organization and regulation of cortical networks

Jekanthan Thangavelautham, Gabriele M T D'Eleuterio

Research output: Contribution to journalArticle

9 Citations (Scopus)

Abstract

A key challenge in evolving control systems for robots using neural networks is training tractability. Evolving monolithic fixed topology neural networks is shown to be intractable with limited supervision in high dimensional search spaces. Common strategies to overcome this limitation are to provide more supervision by encouraging particular solution strategies, manually decomposing the task and segmenting the search space and network. These strategies require a supervisor with domain knowledge and may not be feasible for difficult tasks where novel concepts are required. The alternate strategy is to use self-organized task decomposition to solve difficult tasks with limited supervision. The artificial neural tissue (ANT) approach presented here uses self-organized task decomposition to solve tasks. ANT inspired by neurobiology combines standard neural networks with a novel wireless signaling scheme modeling chemical diffusion of neurotransmitters. These chemicals are used to dynamically activate and inhibit wired network of neurons using a coarse-coding framework. Using only a global fitness function that does not encourage a predefined solution, modular networks of neurons are shown to self-organize and perform task decomposition. This approach solves the sign-following task found to be intractable with conventional fixed and variable topology networks. In this paper, key attributes of the ANT architecture that perform self-organized task decomposition are shown. The architecture is robust and scalable to number of neurons, synaptic connections, and initialization parameters.

Original languageEnglish (US)
Article number6140581
Pages (from-to)552-564
Number of pages13
JournalIEEE Transactions on Neural Networks and Learning Systems
Volume23
Issue number4
DOIs
StatePublished - 2012
Externally publishedYes

Fingerprint

Neurons
Decomposition
Tissue
Neural networks
Topology
Supervisory personnel
Robots
Control systems
Neurobiology

Keywords

  • Coarse coding
  • evolutionary algorithms
  • robotics
  • task decomposition

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Networks and Communications
  • Computer Science Applications
  • Software

Cite this

Tackling learning intractability through topological organization and regulation of cortical networks. / Thangavelautham, Jekanthan; D'Eleuterio, Gabriele M T.

In: IEEE Transactions on Neural Networks and Learning Systems, Vol. 23, No. 4, 6140581, 2012, p. 552-564.

Research output: Contribution to journalArticle

@article{74b7f78824ea4f5a9cdf737a05d04d86,
title = "Tackling learning intractability through topological organization and regulation of cortical networks",
abstract = "A key challenge in evolving control systems for robots using neural networks is training tractability. Evolving monolithic fixed topology neural networks is shown to be intractable with limited supervision in high dimensional search spaces. Common strategies to overcome this limitation are to provide more supervision by encouraging particular solution strategies, manually decomposing the task and segmenting the search space and network. These strategies require a supervisor with domain knowledge and may not be feasible for difficult tasks where novel concepts are required. The alternate strategy is to use self-organized task decomposition to solve difficult tasks with limited supervision. The artificial neural tissue (ANT) approach presented here uses self-organized task decomposition to solve tasks. ANT inspired by neurobiology combines standard neural networks with a novel wireless signaling scheme modeling chemical diffusion of neurotransmitters. These chemicals are used to dynamically activate and inhibit wired network of neurons using a coarse-coding framework. Using only a global fitness function that does not encourage a predefined solution, modular networks of neurons are shown to self-organize and perform task decomposition. This approach solves the sign-following task found to be intractable with conventional fixed and variable topology networks. In this paper, key attributes of the ANT architecture that perform self-organized task decomposition are shown. The architecture is robust and scalable to number of neurons, synaptic connections, and initialization parameters.",
keywords = "Coarse coding, evolutionary algorithms, robotics, task decomposition",
author = "Jekanthan Thangavelautham and D'Eleuterio, {Gabriele M T}",
year = "2012",
doi = "10.1109/TNNLS.2011.2178311",
language = "English (US)",
volume = "23",
pages = "552--564",
journal = "IEEE Transactions on Neural Networks and Learning Systems",
issn = "2162-237X",
publisher = "IEEE Computational Intelligence Society",
number = "4",

}

TY - JOUR

T1 - Tackling learning intractability through topological organization and regulation of cortical networks

AU - Thangavelautham, Jekanthan

AU - D'Eleuterio, Gabriele M T

PY - 2012

Y1 - 2012

N2 - A key challenge in evolving control systems for robots using neural networks is training tractability. Evolving monolithic fixed topology neural networks is shown to be intractable with limited supervision in high dimensional search spaces. Common strategies to overcome this limitation are to provide more supervision by encouraging particular solution strategies, manually decomposing the task and segmenting the search space and network. These strategies require a supervisor with domain knowledge and may not be feasible for difficult tasks where novel concepts are required. The alternate strategy is to use self-organized task decomposition to solve difficult tasks with limited supervision. The artificial neural tissue (ANT) approach presented here uses self-organized task decomposition to solve tasks. ANT inspired by neurobiology combines standard neural networks with a novel wireless signaling scheme modeling chemical diffusion of neurotransmitters. These chemicals are used to dynamically activate and inhibit wired network of neurons using a coarse-coding framework. Using only a global fitness function that does not encourage a predefined solution, modular networks of neurons are shown to self-organize and perform task decomposition. This approach solves the sign-following task found to be intractable with conventional fixed and variable topology networks. In this paper, key attributes of the ANT architecture that perform self-organized task decomposition are shown. The architecture is robust and scalable to number of neurons, synaptic connections, and initialization parameters.

AB - A key challenge in evolving control systems for robots using neural networks is training tractability. Evolving monolithic fixed topology neural networks is shown to be intractable with limited supervision in high dimensional search spaces. Common strategies to overcome this limitation are to provide more supervision by encouraging particular solution strategies, manually decomposing the task and segmenting the search space and network. These strategies require a supervisor with domain knowledge and may not be feasible for difficult tasks where novel concepts are required. The alternate strategy is to use self-organized task decomposition to solve difficult tasks with limited supervision. The artificial neural tissue (ANT) approach presented here uses self-organized task decomposition to solve tasks. ANT inspired by neurobiology combines standard neural networks with a novel wireless signaling scheme modeling chemical diffusion of neurotransmitters. These chemicals are used to dynamically activate and inhibit wired network of neurons using a coarse-coding framework. Using only a global fitness function that does not encourage a predefined solution, modular networks of neurons are shown to self-organize and perform task decomposition. This approach solves the sign-following task found to be intractable with conventional fixed and variable topology networks. In this paper, key attributes of the ANT architecture that perform self-organized task decomposition are shown. The architecture is robust and scalable to number of neurons, synaptic connections, and initialization parameters.

KW - Coarse coding

KW - evolutionary algorithms

KW - robotics

KW - task decomposition

UR - http://www.scopus.com/inward/record.url?scp=84876947700&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84876947700&partnerID=8YFLogxK

U2 - 10.1109/TNNLS.2011.2178311

DO - 10.1109/TNNLS.2011.2178311

M3 - Article

C2 - 24805039

AN - SCOPUS:84876947700

VL - 23

SP - 552

EP - 564

JO - IEEE Transactions on Neural Networks and Learning Systems

JF - IEEE Transactions on Neural Networks and Learning Systems

SN - 2162-237X

IS - 4

M1 - 6140581

ER -