DoB-SNN: A New Neuron Assembly-Inspired Spiking Neural Network for Pattern Classification

Saranirad, Vahid; McGinnity, TM; Dora, Shirin; Coyle, Damien

doi:10.1109/ijcnn52387.2021.9534283

Cited by 2 publications

(4 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Then, all the weights and DoBs in this layer are frozen before training the output layer. The training algorithm for this layer is described in more detail in [22].…”

Section: Scenario Bmentioning

confidence: 99%

“…Three-layer DoB-SNN is compared with two-layer DoB-SNN as well as three other SNNs, including SWAT [9], SRESN [25], and SpikeProp [26] (Table II). The results of two-layer DoB-SNN, SWAT, SRESN, and SpikeProp have been reproduced from [22]. Also, the simulation parameters of three-layer DoB-SNN have been set as in [22].…”

Section: Performance Evaluationmentioning

confidence: 99%

“…The results of two-layer DoB-SNN, SWAT, SRESN, and SpikeProp have been reproduced from [22]. Also, the simulation parameters of three-layer DoB-SNN have been set as in [22]. Three-layer DoB-SNN was implemented in Python 3.7 utilising computing resources provided by the Northern Ireland High-Performance Computing (NI-HPC) facility.…”

Section: Performance Evaluationmentioning

confidence: 99%

“…An assembly of neurons is a group of neurons assigned to a certain piece of cognitive data, such as a memory, a concept, or a phrase [20]. Recently Saranirad et al [22] introduced a new type of spiking neural network inspired by biological assemblies, known as DoB-SNN. All the above SNNs allocate each output neuron to a class before the training process.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Assembly-based STDP: A New Learning Rule for Spiking Neural Networks Inspired by Biological Assemblies

Saranirad

Dora

McGinnity

et al. 2022

2022 International Joint Conference on Neural Networks (IJCNN)

Self Cite

View full text Add to dashboard Cite

Spiking Neural Networks (SNNs), An alternative to sigmoidal neural networks, include time into their operations using discrete signals called spikes. Employing spikes enables SNNs to mimic any feedforward sigmoidal neural network with lower power consumption. Recently a new type of SNN has been introduced for classification problems, known as Degree of Belonging SNN (DoB-SNN). DoB-SNN is a two-layer spiking neural network that shows significant potential as an alternative SNN architecture and learning algorithm. This paper introduces a new variant of Spike-Timing Dependent Plasticity (STDP), which is based on the assembly of neurons and expands the DoB-SNN's training algorithm for multilayer architectures. The new learning rule, known as assembly-based STDP, employs trained DoBs in each layer to train the next layer and build strong connections between neurons from the same assembly while creating inhibitory connections between neurons from different assemblies in two consecutive layers. The performance of the multilayer DoB-SNN is evaluated on five datasets from the UCI machine learning repository. Detailed comparisons on these datasets with other supervised learning algorithms show that the multilayer DoB-SNN can achieve better performance on 4/5 datasets and comparable performance on 5th when compared to multilayer algorithms that employ considerably more trainable parameters.

show abstract

“…Then, all the weights and DoBs in this layer are frozen before training the output layer. The training algorithm for this layer is described in more detail in [22].…”

Section: Scenario Bmentioning

confidence: 99%