Brain-inspired reward broadcasting: Brain learning mechanism guides learning of spiking neural network

Miao Wang, Gangyi Ding, Yunlin Lei, Yu Zhang, Lanyu Gao, Xu Yang

Abstract: Highlights•Treat synapses as intelligent agents and use reinforcement learning to model the brain’s reward mechanism to achieve plasticity.•Two-step learning training method is constructed to enhance the generalization of the neural network.•Implement an SNN brain-like learning algorithm for learning without inter-layer backpropagation.•Verify the benefits of brain-inspired learning in improving neural network performance and generalization.