The Feed-Forward network is a network consisting of one or more neural layers, in which the signal conductors pass only one dimension from the input through layers, up to the output. Each Feed-Forwrad network must have at least one input layout (note that the input layer is not counted as a network layer), a output layer and can have multiple hidden layers. Each neural in the previous class will connect with all the neural in the next layer. The following is a model of the Feed-Forward network:
正在翻譯中..