Hi, I am a student learning about neural networks and the first thing I had to do was to write a hopfield neural networks with asynchronous update then I have to modify that program so that it uses the stochastic update rule and noise but I'm lost a that could someone explain to me what is thata or how to achieve it

thanks

for the discrete case, asynchronous update means neural units do not get updated synchronously. Usually, the update is done by randomly picking a neural unit, which is likely what the stochastic update rule is referring to, but it could also be non-random (for example, by starting with neuron of index 1 and continuing until neuron n (in an n neuron network).

The hopfield Neural Network, is simple it consists of single layer set of neurons called associative memory a hopfield network can store in its memory a sets of patterns for later use, the training algorithm is very simple for a computer implementation, later i will post an url with complete library for training patterns.


Regards