To overcome the problems of the sigmoid function, we will introduce an activation function named hyperbolic tangent function (TanH). The equation of TanH is given in Figure 9.34:
Figure 9.34: Tanh activation function equation (Image credit: https://cdn-images-1.medium.com/max/800/1*HJhu8BO7KxkjqRRMSaz0Gw.png)
This function squashes the input region in the range of [-1 to 1] so its output is zero-centric, which makes optimization easier for us. This function also suffers from the vanishing gradient problem, so we need to see other activation functions.
