Introduction to spiking neural networks: Information processing, learning and applications

Area of Science:

• Neurobiology
• Computational Neuroscience
• Artificial Intelligence

Background:

• The traditional view of neural information encoding relies on neuronal firing rates.
• This firing rate paradigm has influenced artificial neural network theory.
• Emerging evidence highlights the importance of precise action potential timing in neural coding.

Purpose of the Study:

• To introduce spiking neural networks (SNNs) as a new class of neural models.
• To summarize the fundamental properties of spiking neurons and networks.
• To provide an overview of spike-based information coding, synaptic plasticity, and learning.

Main Methods:

• Review of physiological experiments demonstrating spike-based coding.
• Summary of theoretical models of spiking neurons and networks.
• Survey of current applications of spiking neural network models.

Main Results:

• Spiking neural networks offer a biologically plausible alternative to traditional artificial neural networks.
• Spike-based coding emphasizes the timing of action potentials for information representation.
• The paper details models for synaptic plasticity and learning within SNNs.

Conclusions:

• Spiking neural networks represent a significant advancement in understanding neural computation.
• These models are crucial for scientists interested in biologically realistic neural processing.
• The study highlights the potential of SNNs in diverse scientific and technological applications.