Tanh Function
The tanh activation function is just another possible function that can be used as anon-linear activation function between layers of a neural network . It shares a many effects in common with the sigmoid activation function . Unlike a sigmoid function that will collude input values between 0 and 1, the Tanh will collude values between-1 and 1. Analogous to the sigmoid function , one of the intriguing parcels of the tanh function is that the outgrowth of tanh can be expressed in terms of the function itself. When to use which Activation Function in a Neural Network? Specifically , it depends on the problem type and the value range of the anticipated affair . For illustration , to prognosticate values that are larger than 1, tanh or sigmoid aren't suitable to be used in the affair subcaste , rather , ReLU can be used . On the other hand , if the affair values have to be in the range () or