How Neuronal Noises Influence the Spiking Neural Networks's Cognitive Learning Process: A Preliminary Study

Area of Science:

• Neuroscience
• Artificial Intelligence
• Computational Neuroscience

Background:

• The Default Mode Network (DMN) is a key brain network with distinct correlated activities.
• DMNs are known to influence various cognitive functions.
• Spiking Neural Networks (SNNs) are bio-inspired computational models.

Purpose of the Study:

• To investigate the impact of DMNs on the performance of SNNs in image classification tasks.
• To explore the bionic principles in selecting SNN models and parameters.
• To evaluate how DMNs influence SNN structure evolution.

Main Methods:

• Utilized the Leaky Integrate-and-Fire (LIF) neuron model.
• Employed Additive White Gaussian Noise (AWGN) to simulate DMN input.
• Implemented a learning algorithm based on Spike-Timing-Dependent Plasticity (STDP).
• Conducted experiments on two-layer and three-layer SNNs.

Main Results:

• Positive influence of DMNs on classification accuracy in a two-layer SNN.
• DMNs positively affected the structural evolution of a three-layer SNN.
• Experimental results indicate enhanced SNN capabilities with DMN integration.

Conclusions:

• DMNs show potential for improving SNN performance in image classification.
• The bionic approach in SNN modeling and parameter setting is effective.
• Future research can further explore DMN-SNN interactions for advanced AI applications.