Indirect Differentiation of Function for a Network of Biologically Plausible Neurons
Abstract
This paper introduces a new method to model differentiation of biologically plausible neurons, introducing the capability for indirectly defining the characteristics for a network of spiking neurons. Due to its biological plausibility and greater potential for computational power, a spiking neuron model is employed as the basic functional unit in our system. the method for designing the architecture (network design, communication structure, and neuron functionality) for networks of spiking neurons has been purely a manual process. in this paper, we propose a new design for the differentiation of a network of spiking neurons, such that these networks can be indirectly specified, thus enabling a method for the automatic creation of a network for a predetermined function. in this manner, the difficulties associated with the manual creation of these networks are overcome, and opportunity is provided for the utilization of these networks more readily for applications. Thus, this paper provides a new method for indirectly constructing these powerful networks, such as could be easily linked to an evolutionary system or other optimization algorithm. © Springer-Verlag Berlin Heidelberg 2003.
Recommended Citation
A. D. Fischer and C. H. Dagli, "Indirect Differentiation of Function for a Network of Biologically Plausible Neurons," Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 2714, pp. 1089 - 1099, Springer, Jan 2003.
The definitive version is available at https://doi.org/10.1007/3-540-44989-2_130
Department(s)
Engineering Management and Systems Engineering
Keywords and Phrases
Artificial Neural Networks; Computational Intelligence; Evolutionary Algorithms; Evolutionary Neural Networks; Integrate-and-Fire Neuron Models; Neuro-Modeling; Spiking Neural Networks
International Standard Book Number (ISBN)
978-354040408-8
International Standard Serial Number (ISSN)
1611-3349; 0302-9743
Document Type
Article - Journal
Document Version
Citation
File Type
text
Language(s)
English
Rights
© 2024 Springer, All rights reserved.
Publication Date
01 Jan 2003