Implements a network to be trained using backprop through time.
Competitive implements a simple competitive network.
CompetitiveNetwork is a small network encompassing a Competitive group.
Builds an Echo-State Network with options for all valid weight configurations.
A standard feed-forward network, where a succession of neuron groups and synapse groups are organized into layers.
Hopfield is a basic implementation of a discrete Hopfield network.
KwtaNetwork implements a k Winner Take All network.
A Least Mean Squares network.
Implements a simple recurrent network (See, e.g, Elman 1991).
SOM implements a Self-Organizing Map network.
SOMNetwork is a small network encompassing an SOM group.
WinnerTakeAll.The neuron with the highest weighted input in a winner-take-all network takes on an upper value, all other neurons take on the lower value.
Specific implementation of competitive learning.
Main forms of Hopfield update rule.