Spiking Neural Networks with Homeostatic Mechanisms

This project implements various homeostatic mechanisms in a spiking neural network (SNN) to maintain network stability and prevent weight overgrowth.

Homeostatic Mechanisms

• Weight Normalization: Ensures synaptic weights remain balanced.
• Synaptic Scaling: Adjusts synaptic strengths to maintain stability.
• Homeostatic Intrinsic Plasticity (HIP): Dynamically adjusts neuron firing thresholds.
• Lateral Inhibition: Suppresses neighboring neurons to prevent runaway excitation.
• Global Inhibition: Scales down network-wide synaptic strengths when activity exceeds a threshold.

Usage

In the network.py you can acitvate or deactive the homeostatic function by setting them to True or False.

Installation

To run the project, clone this repository and install the required dependencies:

Installation

git clone https://github.com/Edogor/Homeostatic-Methods-Modeling-.git
cd Homeostatic-Methods-Modeling
python -m venv venv
.\venv\Scripts\activate
pip install -r requirements.txt

execution

python network.py