Programming with simulated neurons: a first design pattern

Evans, Carl and Mitchell, Ian and Huyck, Christian R. (2016) Programming with simulated neurons: a first design pattern. In: PPIG 2016 - 27th Annual Workshop of the Psychology of Programming Interest Group, 07-10 Sept 2016, University of Cambridge, Cambridge, UK.

[img]
Preview
PDF - Final accepted version (with author's formatting)
Download (387kB) | Preview

Abstract

An investigation has been carried out with regard to programming a form of deterministic logic based entirely in terms of biologically plausible neurons. To this end, a prototype has been successfully developed that incorporates a neuron version of the classic state design pattern. This neuron version is based on a novel programming technique, which models logical states as persistently active cell assemblies. These are populations of intra-connected neurons that have been triggered to continually fire until programmatically suppressed, thus enabling a neural form of state-transition logic. These neural-state cell assemblies have been developed using a specialist neuron simulation software library that is commonly employed by neuroscientists and is the adopted software protocol for the hardware platforms currently being developed for the Human Brain Project. An underlying inspiration of the work is to look forward to the possibility of a programming paradigm based entirely on biologically plausible neurons. It is envisaged that such a neural programming paradigm would benefit from established techniques, and that the neural cell assembly state pattern that has been developed and described in this report is a next step in that direction. In addition, a new graphical notation has been formulated in order to visualise the prototype. Whilst not a primary focus of the research to date, this visualisation notation may prove beneficial to the computational neuroscience community who work with similar neuron simulation software as that employed for the prototype presented here.

Item Type: Conference or Workshop Item (Paper)
Research Areas: A. > School of Science and Technology > Computer Science > Artificial Intelligence group
Item ID: 20794
Notes on copyright: © 2016, the authors. Permission granted on 26/02/17, by the PPIG (http://www.ppig.org/) to make the full text of the author version available in this Repository (http://eprints.mdx.ac.uk/). The published version will be available at http://www.ppig.org/workshops/ppig-2016-27th-annual-workshop
Useful Links:
Depositing User: Carl Evans
Date Deposited: 25 Oct 2016 09:41
Last Modified: 17 Oct 2017 15:59
URI: http://eprints.mdx.ac.uk/id/eprint/20794

Actions (login required)

Edit Item Edit Item

Full text downloads (NB count will be zero if no full text documents are attached to the record)

Downloads per month over the past year