A brain-inspired algorithm that mitigates catastrophic forgetting of artificial and spiking neural networks with low computational cost

doi:10.1126/sciadv.adi2947

CORC > 自动化研究所 > 中国科学院自动化研究所 > 复杂系统认知与决策实验室

	A brain-inspired algorithm that mitigates catastrophic forgetting of artificial and spiking neural networks with low computational cost
	Zhang, Tielin 1,2,3; Cheng, Xiang 1,2; Jia, Shuncheng 1,2; Li, Chengyu T.3,4; Poo, Mu-ming 3,4; Xu, Bo 1,2
刊名	SCIENCE ADVANCES
	2023-08-01
卷号	9 期号:34 页码:12
ISSN号	2375-2548
DOI	10.1126/sciadv.adi2947
通讯作者	Zhang, Tielin(tielin.zhang@ia.ac.cn) ; Xu, Bo(xubo@ia.ac.cn)
英文摘要	Neuromodulators in the brain act globally at many forms of synaptic plasticity, represented as metaplasticity, which is rarely considered by existing spiking (SNNs) and nonspiking artificial neural networks (ANNs). Here, we report an efficient brain-inspired computing algorithm for SNNs and ANNs, referred to here as neuromodulation-assisted credit assignment (NACA), which uses expectation signals to induce defined levels of neuromodulators to selective synapses, whereby the long-term synaptic potentiation and depression are modified in a nonlinear manner depending on the neuromodulator level. The NACA algorithm achieved high recognition accuracy with substantially reduced computational cost in learning spatial and temporal classification tasks. Notably, NACA was also verified as efficient for learning five different class continuous learning tasks with varying degrees of complexity, exhibiting a markedly mitigated catastrophic forgetting at low computational cost. Mapping synaptic weight changes showed that these benefits could be explained by the sparse and targeted synaptic modifications attributed to expectation-based global neuromodulation.
资助项目	Strategic Priority Research Program of Chinese Academy of Sciences[XDA27010404] ; Shanghai Municipal Science and Technology Major Project[2021SHZDZX] ; Lingang Laboratory Fund[LG202105-01-07] ; Youth Innovation Promotion Association of CAS
WOS关键词	TIMING-DEPENDENT PLASTICITY ; DOPAMINE ; STDP ; SENSITIVITY ; MODULATION ; RECEPTORS
WOS研究方向	Science & Technology - Other Topics
语种	英语
出版者	AMER ASSOC ADVANCEMENT SCIENCE
WOS记录号	WOS:001054596800013
资助机构	Strategic Priority Research Program of Chinese Academy of Sciences ; Shanghai Municipal Science and Technology Major Project ; Lingang Laboratory Fund ; Youth Innovation Promotion Association of CAS
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/54161]
专题	复杂系统认知与决策实验室
通讯作者	Zhang, Tielin; Xu, Bo
作者单位	1.Chinese Acad Sci, Inst Automat, Beijing 100190, Peoples R China 2.Univ Chinese Acad Sci, Sch Artificial Intelligence, Beijing 100049, Peoples R China 3.Shanghai Ctr Brain Sci & Brain inspired Technol, Lingang Lab, Shanghai 200031, Peoples R China 4.Chinese Acad Sci, Inst Neurosci, Ctr Excellence Brain Sci & Intelligence Technol, Shanghai 200031, Peoples R China
推荐引用方式 GB/T 7714	Zhang, Tielin,Cheng, Xiang,Jia, Shuncheng,et al. A brain-inspired algorithm that mitigates catastrophic forgetting of artificial and spiking neural networks with low computational cost[J]. SCIENCE ADVANCES,2023,9(34):12.
APA	Zhang, Tielin,Cheng, Xiang,Jia, Shuncheng,Li, Chengyu T.,Poo, Mu-ming,&Xu, Bo.(2023).A brain-inspired algorithm that mitigates catastrophic forgetting of artificial and spiking neural networks with low computational cost.SCIENCE ADVANCES,9(34),12.
MLA	Zhang, Tielin,et al."A brain-inspired algorithm that mitigates catastrophic forgetting of artificial and spiking neural networks with low computational cost".SCIENCE ADVANCES 9.34(2023):12.