https://github.com/amanpriyanshu/spiking-neural-networks
Spiking Neural Network (SNN). Because brains are the most optimal neural nets, why not accept what has competed for over 3 billion years and still survived.
https://github.com/amanpriyanshu/spiking-neural-networks
Last synced: about 2 months ago
JSON representation
Spiking Neural Network (SNN). Because brains are the most optimal neural nets, why not accept what has competed for over 3 billion years and still survived.
- Host: GitHub
- URL: https://github.com/amanpriyanshu/spiking-neural-networks
- Owner: AmanPriyanshu
- License: mit
- Created: 2021-03-17T03:17:03.000Z (over 4 years ago)
- Default Branch: main
- Last Pushed: 2021-03-17T05:18:43.000Z (over 4 years ago)
- Last Synced: 2025-02-08T17:23:37.875Z (8 months ago)
- Language: Jupyter Notebook
- Size: 176 KB
- Stars: 0
- Watchers: 2
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# Spiking-Neural-Networks
Spiking Neural Network (SNN). Because brains are the most optimal neural nets, why not accept what has competed for over 3 billion years and still survived.
The difference between an ordinary artificial neural network (ANN) and SNN is that, a SNN's neuron only activates/spikes upon crossing a threshold value. The most common model for simulating or exhibiting such spikes are integerate&fire or leaky integerate&fire. More powerful than traditional machine learning techniques, it is rare to see such powerful implementation not widespread in the booming ML/AI world. However, an important reason for such an effect, is the inability to differentiate such spiking or binary functions. In essence, it is our incapacity to train in a supervised fashion such a model with efficient learning progress. Therefore, it has seen rare use cases, and can only be trained using unsupervised learning methods.