A spiking neural learning classifier system

Howard, G. D., Bull, L. and Lanzi, P. L. (2012) A spiking neural learning classifier system. arXiv preprint arXiv:1201.3249. Available from: http://eprints.uwe.ac.uk/20727

Full text not available from this repository

Publisher's URL: http://arxiv.org


Learning Classifier Systems (LCS) are population-based reinforcement learners used in a wide variety of applications. This paper presents a LCS where each traditional rule is represented by a spiking neural network, a type of network with dynamic internal state. We employ a constructivist model of growth of both neurons and dendrites that realise flexible learning by evolving structures of sufficient complexity to solve a well-known problem involving continuous, real-valued inputs. Additionally, we extend the system to enable temporal state decomposition. By allowing our LCS to chain together sequences of heterogeneous actions into macro-actions, it is shown to perform optimally in a problem where traditional methods can fail to find a solution in a reasonable amount of time. Our final system is tested on a simulated robotics platform.

Item Type:Article
Uncontrolled Keywords:robotics, learning, computer science, neural and evolutionary computing
Faculty/Department:Faculty of Environment and Technology > Department of Computer Science and Creative Technologies
ID Code:20727
Deposited By: M. Clarke
Deposited On:12 Jul 2013 15:59
Last Modified:17 Apr 2016 18:46

Request a change to this item