AIFH Volume 3, Chapter 3: Hopfield Network

This page allows you to train and run a Hopfield network. The weights that result from this training are shown below.

Training Data

Training pattern:

01

01

01

01

Weights

The weight matrix determines the connections between neurons. No neuron is self-connected, so the diagonal should always be zero. You can manually enter weights, however, they must be numeric.

	0	1	2	3
0
1
2
3

Run

Individual neurons can be run, their state will be updated based on the weights and other states. You can also run all neurons.

Current state:

01	01	01	01