Sensitivity of Spiking Neural Networks Due to Input Perturbation
<b>Background:</b> To investigate the behavior of spiking neural networks (SNNs), the sensitivity of input perturbation serves as an effective metric for assessing the influence on the network output. However, existing methods fall short in evaluating the sensitivity of SNNs featuring bi...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2024-11-01
|
| Series: | Brain Sciences |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2076-3425/14/11/1149 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1846154128364929024 |
|---|---|
| author | Haoran Zhu Xiaoqin Zeng Yang Zou Jinfeng Zhou |
| author_facet | Haoran Zhu Xiaoqin Zeng Yang Zou Jinfeng Zhou |
| author_sort | Haoran Zhu |
| collection | DOAJ |
| description | <b>Background:</b> To investigate the behavior of spiking neural networks (SNNs), the sensitivity of input perturbation serves as an effective metric for assessing the influence on the network output. However, existing methods fall short in evaluating the sensitivity of SNNs featuring biologically plausible leaky integrate-and-fire (LIF) neurons due to the intricate neuronal dynamics during the feedforward process. <b>Methods:</b> This paper first defines the sensitivity of a temporal-coded spiking neuron (SN) as the deviation between the perturbed and unperturbed output under a given input perturbation with respect to overall inputs. Then, the sensitivity algorithm of an entire SNN is derived iteratively from the sensitivity of each individual neuron. Instead of using the actual firing time, the desired firing time is employed to derive a more precise analytical expression of the sensitivity. Moreover, the expectation of the membrane potential difference is utilized to quantify the magnitude of the input deviation. <b>Results/Conclusions:</b> The theoretical results achieved with the proposed algorithm are in reasonable agreement with the simulation results obtained with extensive input data. The sensitivity also varies monotonically with changes in other parameters, except for the number of time steps, providing valuable insights for choosing appropriate values to construct the network. Nevertheless, the sensitivity exhibits a piecewise decreasing tendency with respect to the number of time steps, with the length and starting point of each piece contingent upon the specific parameter values of the neuron. |
| format | Article |
| id | doaj-art-d1e9ef87b7ff4aea95c02bcbef1ecf15 |
| institution | Kabale University |
| issn | 2076-3425 |
| language | English |
| publishDate | 2024-11-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Brain Sciences |
| spelling | doaj-art-d1e9ef87b7ff4aea95c02bcbef1ecf152024-11-26T17:55:11ZengMDPI AGBrain Sciences2076-34252024-11-011411114910.3390/brainsci14111149Sensitivity of Spiking Neural Networks Due to Input PerturbationHaoran Zhu0Xiaoqin Zeng1Yang Zou2Jinfeng Zhou3College of Computer Science and Software Engineering, Hohai University, Nanjing 211100, ChinaCollege of Computer Science and Software Engineering, Hohai University, Nanjing 211100, ChinaCollege of Computer Science and Software Engineering, Hohai University, Nanjing 211100, ChinaCollege of Computer Science and Software Engineering, Hohai University, Nanjing 211100, China<b>Background:</b> To investigate the behavior of spiking neural networks (SNNs), the sensitivity of input perturbation serves as an effective metric for assessing the influence on the network output. However, existing methods fall short in evaluating the sensitivity of SNNs featuring biologically plausible leaky integrate-and-fire (LIF) neurons due to the intricate neuronal dynamics during the feedforward process. <b>Methods:</b> This paper first defines the sensitivity of a temporal-coded spiking neuron (SN) as the deviation between the perturbed and unperturbed output under a given input perturbation with respect to overall inputs. Then, the sensitivity algorithm of an entire SNN is derived iteratively from the sensitivity of each individual neuron. Instead of using the actual firing time, the desired firing time is employed to derive a more precise analytical expression of the sensitivity. Moreover, the expectation of the membrane potential difference is utilized to quantify the magnitude of the input deviation. <b>Results/Conclusions:</b> The theoretical results achieved with the proposed algorithm are in reasonable agreement with the simulation results obtained with extensive input data. The sensitivity also varies monotonically with changes in other parameters, except for the number of time steps, providing valuable insights for choosing appropriate values to construct the network. Nevertheless, the sensitivity exhibits a piecewise decreasing tendency with respect to the number of time steps, with the length and starting point of each piece contingent upon the specific parameter values of the neuron.https://www.mdpi.com/2076-3425/14/11/1149spiking neural networksensitivitytemporal codingleaky integrate-and-fire |
| spellingShingle | Haoran Zhu Xiaoqin Zeng Yang Zou Jinfeng Zhou Sensitivity of Spiking Neural Networks Due to Input Perturbation Brain Sciences spiking neural network sensitivity temporal coding leaky integrate-and-fire |
| title | Sensitivity of Spiking Neural Networks Due to Input Perturbation |
| title_full | Sensitivity of Spiking Neural Networks Due to Input Perturbation |
| title_fullStr | Sensitivity of Spiking Neural Networks Due to Input Perturbation |
| title_full_unstemmed | Sensitivity of Spiking Neural Networks Due to Input Perturbation |
| title_short | Sensitivity of Spiking Neural Networks Due to Input Perturbation |
| title_sort | sensitivity of spiking neural networks due to input perturbation |
| topic | spiking neural network sensitivity temporal coding leaky integrate-and-fire |
| url | https://www.mdpi.com/2076-3425/14/11/1149 |
| work_keys_str_mv | AT haoranzhu sensitivityofspikingneuralnetworksduetoinputperturbation AT xiaoqinzeng sensitivityofspikingneuralnetworksduetoinputperturbation AT yangzou sensitivityofspikingneuralnetworksduetoinputperturbation AT jinfengzhou sensitivityofspikingneuralnetworksduetoinputperturbation |