Towards a learning-theoretic analysis of spike-timing dependent plasticity

TL;DR

This paper introduces the selectron model, a theoretical framework for analyzing spike-timing dependent plasticity (STDP), demonstrating how it encodes reward signals and proposing regularization to enhance learning robustness.

Contribution

The paper presents the selectron model as a tractable, theoretical abstraction of STDP, linking spike activity to reward estimation and introducing a regularized STDP for improved robustness.

Findings

Selectron encodes reward estimates into spikes.

Error bounds on spikes depend on spiking margin and synaptic weights.

Regularized STDP enhances learning robustness with multiple stimuli.

Abstract

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.