Characterization of Generalizability of Spike Timing Dependent Plasticity trained Spiking Neural Networks

TL;DR

This paper investigates how Spike Timing Dependent Plasticity (STDP) influences the generalizability of Spiking Neural Networks by analyzing learning trajectories and optimizing hyper-parameters through Bayesian methods.

Contribution

It introduces a novel analysis of STDP-based SNNs using Hausdorff dimension and develops a Bayesian optimization approach for hyper-parameter tuning to enhance generalizability.

Findings

Hausdorff dimension characterizes SNN generalizability

Hyper-parameters significantly affect learning trajectory complexity

Bayesian optimization improves SNN generalization performance

Abstract

A Spiking Neural Network (SNN) is trained with Spike Timing Dependent Plasticity (STDP), which is a neuro-inspired unsupervised learning method for various machine learning applications. This paper studies the generalizability properties of the STDP learning processes using the Hausdorff dimension of the trajectories of the learning algorithm. The paper analyzes the effects of STDP learning models and associated hyper-parameters on the generalizability properties of an SNN. The analysis is used to develop a Bayesian optimization approach to optimize the hyper-parameters for an STDP model for improving the generalizability properties of an SNN.