(Solved):   Which of the following is TRUE regarding the nonlinear activation functions (such as Sigmo ...

 

Which of the following is TRUE regarding the nonlinear activation functions (such as Sigmoid, ReLU, tanh) within neural networks? They help to learn nonlinear decision boundaries. They can speed up the gradient calculation in backpropagation, as compared to linear units. They are applied only to the hidden units. They must always output positive values.