Backprop Network

A backprop network is a feed forward network that can be trained using the backpropagation learning algorithm, which is perhaps the best known and most popular means of training neural networks.

To use backprop networks, they must be created using Insert network > backprop, trained, and then used.

Note that backprop can also be used separately from a backprop network, in scripts to custom train a set of weights.

An example which steps through the process of creating and trainining a backprop network is in the examples page.

Creation Dialog

The backprop creation dialog allows you to specify the network topology and neuron types. This dialog is documented in feed forward docs.

Training

Training a network involves specifying input data, target data, and then running the algorithm. This process is covered here.

Once the network is trained to perform a particular input-output mapping, it can be linked to other networks or neurons in Simbrain, or simply used on its own, primarily with the validation tab.