Neural network IP-Core

Limitations:

• Number of neurons should be a power of two

• Number of neurons in each hidden layer are equal

• Activation function of all hidden layer are equal

• Output layer activation function is independent of hidden layer activation function