Phenomenological models of synaptic plasticity based on spike timing

Area of Science:

• Computational Neuroscience
• Neuroscience

Background:

• Synaptic plasticity is the fundamental mechanism underlying learning and memory.
• Phenomenological models offer simplified yet powerful representations of synaptic changes.

Purpose of the Study:

• To establish a framework for classifying and evaluating phenomenological models of synaptic plasticity.
• To focus on models compatible with integrate-and-fire neuron models for large-scale simulations.

Main Methods:

• Reviewing phenomenological models of short-term and long-term synaptic plasticity.
• Analyzing models based on spike timing, membrane potential, and synaptic weight.
• Evaluating model compatibility with experimental data and theoretical expectations.

Main Results:

• Identified update rules dependent on spike timing and filtered variables.
• Demonstrated suitability of discussed models for large-scale network simulations.
• Explored connections between synaptic plasticity models and supervised/reinforcement learning.

Conclusions:

• Phenomenological models provide a viable framework for studying synaptic plasticity.
• These models are essential for understanding learning and memory at a network level.
• The reviewed models support computational neuroscience research and large-scale simulations.