Towards a learning-theoretic analysis of spike-timing dependent plasticity
This work addresses a foundational problem in computational neuroscience by providing a learning-theoretic analysis of spike-timing dependent plasticity, offering incremental theoretical insights for researchers in neural modeling and brain-inspired AI.
The paper tackles the problem of understanding how synaptic plasticity benefits brain function by introducing a theoretical model called the selectron, derived from neurons with spike-timing dependent plasticity, and shows that it encodes reward estimates into spikes with error bounds controlled by spiking margin and synaptic weights, proposing a regularized STDP that improves robustness in multi-stimulus learning.
This paper suggests a learning-theoretic perspective on how synaptic plasticity benefits global brain functioning. We introduce a model, the selectron, that (i) arises as the fast time constant limit of leaky integrate-and-fire neurons equipped with spiking timing dependent plasticity (STDP) and (ii) is amenable to theoretical analysis. We show that the selectron encodes reward estimates into spikes and that an error bound on spikes is controlled by a spiking margin and the sum of synaptic weights. Moreover, the efficacy of spikes (their usefulness to other reward maximizing selectrons) also depends on total synaptic strength. Finally, based on our analysis, we propose a regularized version of STDP, and show the regularization improves the robustness of neuronal learning when faced with multiple stimuli.