Low-error approximation of artificial neuron sigmoid function and its derivative

A new low-error approximation of the sigmoid function based on the piecewise linear method is proposed. The approximation results, in comparison with those of the state-of-the-art, show the lowest mean absolute and relative errors.