For the first question. After you have trained the model and have a 1x1 struct in your workspace. You can save it in a mat file and load it again when you want to use it. No need to train each time you need to use it,
save('filename.mat', 'variableName');
to load the data back
data = load('filename.mat')
you trained model will be inside struct data.
For your second question. If you have already trained the dataset, the training algorithm has already determined the weights based on the training dataset. You don't need to do it yourself.
