{"title":"通过增强直接反馈排列训练尖峰神经网络","authors":"Yongbo Zhang, Katsuma Inoue, Mitsumasa Nakajima, Toshikazu Hashimoto, Yasuo Kuniyoshi, Kohei Nakajima","doi":"arxiv-2409.07776","DOIUrl":null,"url":null,"abstract":"Spiking neural networks (SNNs), the models inspired by the mechanisms of real\nneurons in the brain, transmit and represent information by employing discrete\naction potentials or spikes. The sparse, asynchronous properties of information\nprocessing make SNNs highly energy efficient, leading to SNNs being promising\nsolutions for implementing neural networks in neuromorphic devices. However,\nthe nondifferentiable nature of SNN neurons makes it a challenge to train them.\nThe current training methods of SNNs that are based on error backpropagation\n(BP) and precisely designing surrogate gradient are difficult to implement and\nbiologically implausible, hindering the implementation of SNNs on neuromorphic\ndevices. Thus, it is important to train SNNs with a method that is both\nphysically implementatable and biologically plausible. In this paper, we\npropose using augmented direct feedback alignment (aDFA), a gradient-free\napproach based on random projection, to train SNNs. This method requires only\npartial information of the forward process during training, so it is easy to\nimplement and biologically plausible. We systematically demonstrate the\nfeasibility of the proposed aDFA-SNNs scheme, propose its effective working\nrange, and analyze its well-performing settings by employing genetic algorithm.\nWe also analyze the impact of crucial features of SNNs on the scheme, thus\ndemonstrating its superiority and stability over BP and conventional direct\nfeedback alignment. Our scheme can achieve competitive performance without\naccurate prior knowledge about the utilized system, thus providing a valuable\nreference for physically training SNNs.","PeriodicalId":501347,"journal":{"name":"arXiv - CS - Neural and Evolutionary Computing","volume":"37 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Training Spiking Neural Networks via Augmented Direct Feedback Alignment\",\"authors\":\"Yongbo Zhang, Katsuma Inoue, Mitsumasa Nakajima, Toshikazu Hashimoto, Yasuo Kuniyoshi, Kohei Nakajima\",\"doi\":\"arxiv-2409.07776\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Spiking neural networks (SNNs), the models inspired by the mechanisms of real\\nneurons in the brain, transmit and represent information by employing discrete\\naction potentials or spikes. The sparse, asynchronous properties of information\\nprocessing make SNNs highly energy efficient, leading to SNNs being promising\\nsolutions for implementing neural networks in neuromorphic devices. However,\\nthe nondifferentiable nature of SNN neurons makes it a challenge to train them.\\nThe current training methods of SNNs that are based on error backpropagation\\n(BP) and precisely designing surrogate gradient are difficult to implement and\\nbiologically implausible, hindering the implementation of SNNs on neuromorphic\\ndevices. Thus, it is important to train SNNs with a method that is both\\nphysically implementatable and biologically plausible. In this paper, we\\npropose using augmented direct feedback alignment (aDFA), a gradient-free\\napproach based on random projection, to train SNNs. This method requires only\\npartial information of the forward process during training, so it is easy to\\nimplement and biologically plausible. We systematically demonstrate the\\nfeasibility of the proposed aDFA-SNNs scheme, propose its effective working\\nrange, and analyze its well-performing settings by employing genetic algorithm.\\nWe also analyze the impact of crucial features of SNNs on the scheme, thus\\ndemonstrating its superiority and stability over BP and conventional direct\\nfeedback alignment. Our scheme can achieve competitive performance without\\naccurate prior knowledge about the utilized system, thus providing a valuable\\nreference for physically training SNNs.\",\"PeriodicalId\":501347,\"journal\":{\"name\":\"arXiv - CS - Neural and Evolutionary Computing\",\"volume\":\"37 1\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-09-12\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - CS - Neural and Evolutionary Computing\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2409.07776\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - CS - Neural and Evolutionary Computing","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2409.07776","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Training Spiking Neural Networks via Augmented Direct Feedback Alignment
Spiking neural networks (SNNs), the models inspired by the mechanisms of real
neurons in the brain, transmit and represent information by employing discrete
action potentials or spikes. The sparse, asynchronous properties of information
processing make SNNs highly energy efficient, leading to SNNs being promising
solutions for implementing neural networks in neuromorphic devices. However,
the nondifferentiable nature of SNN neurons makes it a challenge to train them.
The current training methods of SNNs that are based on error backpropagation
(BP) and precisely designing surrogate gradient are difficult to implement and
biologically implausible, hindering the implementation of SNNs on neuromorphic
devices. Thus, it is important to train SNNs with a method that is both
physically implementatable and biologically plausible. In this paper, we
propose using augmented direct feedback alignment (aDFA), a gradient-free
approach based on random projection, to train SNNs. This method requires only
partial information of the forward process during training, so it is easy to
implement and biologically plausible. We systematically demonstrate the
feasibility of the proposed aDFA-SNNs scheme, propose its effective working
range, and analyze its well-performing settings by employing genetic algorithm.
We also analyze the impact of crucial features of SNNs on the scheme, thus
demonstrating its superiority and stability over BP and conventional direct
feedback alignment. Our scheme can achieve competitive performance without
accurate prior knowledge about the utilized system, thus providing a valuable
reference for physically training SNNs.