TY  - GEN
CY  - Newcastle, United Kingdom
A1  - Kozdon, K
A1  - Bentley, P
KW  - spiking neural networks
KW  -  neural networks
KW  -  AI
N2  - Maintaining the ability to fire sparsely is crucial for infor- mation encoding in neural networks. Additionally, spiking homeostasis is vital for spiking neural networks with chang- ing numbers of weights and neurons. We discuss a range of network stabilisation approaches, inspired by homeostatic synaptic plasticity mechanisms reported in the brain. These include weight scaling, and weight change as a function of the network?s spiking activity. We tested normalisation of the sum of weights for all neurons, and by neuron type. We ex- amined how this approach affects firing rate and performance on clustering of time-series data in the form of moving geo- metric shapes. We found that neuron type-specific normali- sation is a promising approach for preventing weight drift in spiking neural networks, thus enabling longer training cycles. It can be adapted for networks with architectural plasticity.
ID  - discovery10080543
PB  - Developmental Neural Networks
UR  - https://www.irit.fr/devonn/files/alife2019-kozdon.pdf
N1  - This version is the version of record. For information on re-use, please refer to the publisher?s terms and conditions.
TI  - Normalisation of Weights and Firing Rates in Spiking Neural Networks with Spike-Timing-Dependent Plasticity
AV  - public
Y1  - 2019/07/29/
ER  -