/normad

Python2 / Brian2 implementation (attempt) of Normalized Approximate Descent based Supervised Learning Rule.

Primary LanguagePython

normad

Python2 / Brian2 implementation (attempt) of Normalized Approximate Descent based Supervised Learning Rule.

A two layer spiking neural network (leaky integrate and fire neurons) using the ReSuMe supervised learning rule. The network reads images from the MNIST database and attempts to classify them. Development is still in progress.

Initially the normad supervised learning rule was also implemented for this, and that may be used later.