Comparative Analysis of Spiking Neurons Mathematical Models Training using Surrogate Gradients Techniques

Conference paper

Abdelkader Haddag, Elisa Guerrero Vazquez, Hayat Yedjour, Maria de La Paz Guerrero Lebrero
Lecture Notes in Computer Science (LNCS), Ignacio Rojas, Gonzalo Joya, Andreu Catala, Advances in Computational Intelligence, vol. 16009, Springer Nature, 2025 Oct 1, pp. 244–257

DOI: 10.1007/978-3-032-02728-3_20

Cite

APA Click to copy
Haddag, A., Vazquez, E. G., Yedjour, H., & de La Paz Guerrero Lebrero, M. (2025). Comparative Analysis of Spiking Neurons Mathematical Models Training using Surrogate Gradients Techniques. In I. Rojas, G. Joya, & A. Catala (Eds.), Advances in Computational Intelligence (Vol. 16009, pp. 244–257). Springer Nature. https://doi.org/10.1007/978-3-032-02728-3_20

Chicago/Turabian Click to copy
Haddag, Abdelkader, Elisa Guerrero Vazquez, Hayat Yedjour, and Maria de La Paz Guerrero Lebrero. “Comparative Analysis of Spiking Neurons Mathematical Models Training Using Surrogate Gradients Techniques.” In Advances in Computational Intelligence, edited by Ignacio Rojas, Gonzalo Joya, and Andreu Catala, 16009:244–257. Lecture Notes in Computer Science (LNCS). Springer Nature, 2025.

MLA Click to copy
Haddag, Abdelkader, et al. “Comparative Analysis of Spiking Neurons Mathematical Models Training Using Surrogate Gradients Techniques.” Advances in Computational Intelligence, edited by Ignacio Rojas et al., vol. 16009, Springer Nature, 2025, pp. 244–57, doi:10.1007/978-3-032-02728-3_20.

BibTeX Click to copy

@inproceedings{abdelkader2025a,
  title = {Comparative Analysis of Spiking Neurons Mathematical Models Training using Surrogate Gradients Techniques},
  year = {2025},
  month = oct,
  day = {1},
  pages = {244–257},
  publisher = {Springer Nature},
  series = {Lecture Notes in Computer Science (LNCS)},
  volume = {16009},
  doi = {10.1007/978-3-032-02728-3_20},
  author = {Haddag, Abdelkader and Vazquez, Elisa Guerrero and Yedjour, Hayat and de La Paz Guerrero Lebrero, Maria},
  editor = {Rojas, Ignacio and Joya, Gonzalo and Catala, Andreu},
  booktitle = {Advances in Computational Intelligence},
  month_numeric = {10}
}

Abstract

Multimodal data is emerging from different sources and in large quantities, which has allowed researchers to train highly-performing intelligent models and agents. However, the computational and environmental costs of the current deep learning trends are against the sustainability goals set worldwide, including the environmental concerns about its carbon footprint. Spiking neural networks offer a bio-plausible alternative given their low computational needs, hence less carbon emissions. In this work, we analyze the performance of different spiking neurons, propose a new spiking layer implementation trained using surrogate gradients, and test against different feature extraction scenarios in fully connected spiking neural networks. We also introduce a new metric to compare in-model and cross-model for better decisions when designing and training spiking neural networks.