A chaotic spiking backpropagation approach to brain learning  

作　　者：Guanrong Chen 

机构地区：[1]Department of Electrical Engineering,City University of Hong Kong,China

出　　处：《National Science Review》2024年第6期76-77,共2页国家科学评论（英文版）

摘　　要：Spiking neural networks(SNNs)[1]are known for their superior energy efficiency,achieved through spiking signal transmission that mirrors biological nervous systems,thus solving the unsustainable energy-consumption problem of current artificial neural networks and frameworks.However,effectively training SNNs poses a significant technical challenge.While surrogate gradient-based methods provide a viable solution,the trained SNNs often get stuck in local minima due to their reliance on the inherent gradient dynamics.

关 键 词：NETWORKS NEURAL CHAOTIC 

分 类 号：R741[医药卫生—神经病学与精神病学]