As for the Backpropagation algorithm, mark the correct alternative considering true (T) or false (F).

( ) Its training process consists of three-step iteration: forward, backward, and upward.

( ) For the algorithm to be used, the activation function needs to be discrete, non-differentiable and, preferably, increasing.

( ) The first step (forward) aims to obtain an output for each neuron in the network, being similar to the process of making a prediction.

ovo In the second step (backward), the calculated errors of the output layers are used to adjust the input weights, proceeding from the output layer to the first hidden layer.

A.

F, T, T, F. B.

F, F, T, T. C.

T, T, T, T. D.

F, T, F, T. E.

T, F, F, T.