Date Thesis Awarded
4-2024
Access Type
Honors Thesis -- Access Restricted On-Campus Only
Degree Name
Bachelors of Science (BS)
Department
Neuroscience
Advisor
Greg Conradi Smith
Committee Members
Christopher A. Del Negro
Junping Shi
Robert C. Barnet
Abstract
Synaptic plasticity is thought to be the biological foundation of learning and memory, and it may also contribute to other brain computations and functions. Spike timing-dependent plasticity (STDP) modulates long-term synaptic strengths, aligning with Hebbian learning principles such as the asymmetric impact of precise temporal correlations between the spikes of pre- and post-synaptic neurons. In the context of an all-to-all connected excitatory network, we investigated STDP through mathematical analysis and computational simulations, incorporating a pair-based STDP model and the homeostatic phenomenon known as synaptic scaling. Neuron spiking is modeled as a time-inhomogeneous Poisson point process with rate given by a nonlinear activity function, an approach that follows a recently published spiking network model of the brainstem preBötzinger complex (preBötC) that is responsible for inspiratory breathing rhythm. This model formulation is used to examine how different STDP rules (e.g., additive vs. multiplicative update) influence the activity and connectivity of an excitatory neuron population. Numerical simulations revealed that STDP dynamics, combined with synaptic scaling, can induce episodic bursting (i.e., periodic transitions between network states with low and high firing rate). This finding raises the possibility that STDP may play a role in recurrent excitatory networks, for example, the inspiratory rhythm of the preBötC.
Recommended Citation
Chen, Yinuo, "The Influence of Spike Timing-Dependent Plasticity on a Recurrent Excitatory Neural Network Model" (2024). Undergraduate Honors Theses. William & Mary. Paper 2238.
https://scholarworks.wm.edu/honorstheses/2238
