Consider training a neural network with a sufficiently small training dataset, in one iteration of the algorithm $($that is$,$ one iteration of parameter update$),$ which is the CORRECT sequence of steps?

$$a$)$Back propagation; Forward propagation: Weight update with gradient descent

$$b$)$Forward propagation; Weight update with gradient descent; Back propagation

$$c$)$Weight update with gradient descent; Forward propagation; Back propagation

$$d$)$Forward propagation: Back propagation: Weight update with gradient descent