Quasi Biologically Plausible Category Learning
Conference paper
Huyck, C. 2024. Quasi Biologically Plausible Category Learning. 44th SGAI International Conference on Artificial Intelligence, AI 2024. Cambridge, UK 17 - 19 Dec 2024 Springer.
Type | Conference paper |
---|---|
Title | Quasi Biologically Plausible Category Learning |
Authors | Huyck, C. |
Abstract | This paper explores machine learning using adaptive spiking neurons and spike timing dependent plasticity (STDP). This is shown to work on two categorisation tasks. It is neuro-biologically flawed but works with a small number of point neurons, and is much closer to biology than multi layer perceptrons. The work is derived from mathematical exploration and the portion of the parameter space where categorisation works is small. This is just a proof of concept that categorisation can be done by these spiking competitive nets with STDP. The parameter space could be further explored to find better results, or how to apply this to new categorisation tasks. This work provides support for further exploration of neurobiologically plausible category learning. |
Keywords | Spiking Neurons; Spike Timing Dependent Plasticity; Categorisation; MNIST |
Sustainable Development Goals | 9 Industry, innovation and infrastructure |
Middlesex University Theme | Creativity, Culture & Enterprise |
Research Group | Artificial Intelligence group |
Conference | 44th SGAI International Conference on Artificial Intelligence, AI 2024 |
Proceedings Title | Artificial Intelligence XLI: 44th SGAI International Conference on Artificial Intelligence, AI 2024, Cambridge, UK, December 17–19, 2024, Proceedings |
Series | Lecture Notes in Computer Science |
ISSN | 0302-9743 |
Electronic | 1611-3349 |
Publisher | Springer |
Publication dates | |
17 Dec 2024 | |
Publication process dates | |
Accepted | 30 Aug 2024 |
Deposited | 25 Sep 2024 |
Output status | Accepted |
Accepted author manuscript | File Access Level Open |
Copyright Statement | This version of the contribution has been accepted for publication, after peer review (when applicable) and is subject to Springer Nature’s AM terms of use (https://www.springernature.com/gp/open-research/policies/accepted-ma...), but is not the Version of Record and does not reflect post-acceptance improvements, or any corrections. The Version of Record is available online at: |
https://repository.mdx.ac.uk/item/19vv1w
Restricted files
Accepted author manuscript
6
total views5
total downloads0
views this month0
downloads this month