To spike or not to spike: a probabilistic spiking neuron model

Area of Science:

• Computational neuroscience
• Artificial intelligence
• Machine learning

Background:

• Spiking Neural Networks (SNNs) offer advanced information processing using spike trains.
• Current deterministic SNN models limit applications in stochastic processes.
• There is a need for probabilistic models to capture complex dynamics.

Purpose of the Study:

• To propose a novel probabilistic spiking neuron model (pSNM).
• To demonstrate the construction of probabilistic Spiking Neural Networks (pSNNs).
• To extend existing computational neurogenetic models.

Main Methods:

• Development of a new probabilistic spiking neuron model (pSNM).
• Design principles for building probabilistic Spiking Neural Networks (pSNNs).
• Integration with computational neurogenetic models.

Main Results:

• The proposed pSNM enables probabilistic information processing in SNNs.
• pSNNs can be effectively constructed for various applications.
• The model extends capabilities for cognitive and engineering tasks.

Conclusions:

• The novel pSNM overcomes limitations of deterministic SNNs.
• pSNNs offer a powerful framework for modeling stochastic processes.
• This work advances SNNs for classification, pattern recognition, and memory applications.