Learning-rate-dependent clustering and self-development in a network of coupled phase oscillators

Area of Science:

• Computational neuroscience
• Complex systems dynamics

Background:

• The Kuramoto model describes coupled phase oscillators.
• Hebbian learning strengthens connections between coactive neurons.

Purpose of the Study:

• Investigate the impact of learning rate on coupled oscillator networks with Hebbian learning.
• Explore self-development of neuronal networks through synchronization and Hebbian learning.

Main Methods:

• Simulated Kuramoto model with dynamically varying Hebbian coupling coefficients.
• Analyzed network behavior across different learning rates.

Main Results:

• Fast learning rates (> critical value) yield two antiphase synchronized clusters and all-to-all coupling.
• Slow learning rates (< critical value) result in a single synchronized cluster.
• Network self-development is hindered by excessively slow learning.

Conclusions:

• Learning rate critically influences network structure and synchronization in Hebbian-coupled oscillators.
• The network exhibits memory-like capabilities, generating and retaining stable patterns.