Hi
trainscg does not directly use an explicit ‘learning rate’ parameter as in case with train functions like traingdx which employees’ gradient descent with learning rate and momentum. The parameters and the available default set of values can be found at the attached documentation.
Also, weights & biases and activation functions are the property of the network and not of the train function used to train them.
You can refer the following page for reference.
Hope this helps.
