Understanding feedforward neural networks
The processing from the input layer to the hidden layer(s) and then to the output layer is called feedforward propagation. The transfer function is applied at each hidden layer, and then the activation function value is propagated to the next layer. The next layer can be another hidden layer or the output layer.
Important note
The term "feedforward" is used to indicate the networks in which each node receives connections only from the lower layers. These networks emit a response for each input pattern but fail to capture the possible temporal structure of the input information or to exhibit endogenous temporal dynamics.
Exploring neural network training
The learning ability of an ANN is manifested in the training procedure. This represents the crucial phase of the whole algorithm as it is through the characteristics extracted during training that the network acquires the ability to generalize. The training takes place...
