TY - GEN
T1 - Spike-based learning of transfer functions with the SpiNNaker neuromimetic simulator
AU - Davies, Sergio
AU - Stewart, Terry
AU - Eliasmith, Chris
AU - Furber, Steve
N1 - The research presented here was initiated during the TellurideNeuromorphic Engineering Workshop 2012, and extendedbeyond this workshop. The authors would like to thankall the organizers.The SpiNNaker project is supported by the Engineeringand Physical Science Research Council (EPSRC), grantEP/4015740/1, and also by ARM and Silistix. We appreciatethe support of these sponsors and industrial partners.The authors would like to thank Dr. Simon Davidson andDr. John Viv Woods for the time dedicated to discussions andreviews of this paper.
PY - 2013
Y1 - 2013
N2 - Recent papers have shown the possibility to implement large scale neural network models that perform complex algorithms in a biologically realistic way. However, such models have been simulated on architectures unable to perform realtime simulations. In previous work we presented the possibility to simulate simple models in real-time on the SpiNNaker neu-romimetic architecture. However, such models were 'static': the algorithm performed was defined at design-time. In this paper we present a novel learning rule, that exploits the peculiarities of the SpiNNaker system, enabling models designed with the Neural Engineering Framework (NEF) to learn transfer functions using a supervised framework. We show that the proposed learning rule, belonging to the Prescribed Error Sensitivity (PES) class, is able to learn, effectively, both linear and non-linear functions. © 2013 IEEE.
AB - Recent papers have shown the possibility to implement large scale neural network models that perform complex algorithms in a biologically realistic way. However, such models have been simulated on architectures unable to perform realtime simulations. In previous work we presented the possibility to simulate simple models in real-time on the SpiNNaker neu-romimetic architecture. However, such models were 'static': the algorithm performed was defined at design-time. In this paper we present a novel learning rule, that exploits the peculiarities of the SpiNNaker system, enabling models designed with the Neural Engineering Framework (NEF) to learn transfer functions using a supervised framework. We show that the proposed learning rule, belonging to the Prescribed Error Sensitivity (PES) class, is able to learn, effectively, both linear and non-linear functions. © 2013 IEEE.
U2 - 10.1109/IJCNN.2013.6706962
DO - 10.1109/IJCNN.2013.6706962
M3 - Conference contribution
SN - 9781467361293
BT - Proceedings of the International Joint Conference on Neural Networks|Proc Int Jt Conf Neural Networks
PB - IEEE
CY - USA
T2 - 2013 International Joint Conference on Neural Networks, IJCNN 2013
Y2 - 1 July 2013
ER -