The Gradient Descent algorithm includes a sequence of steps. Which is the correct ordering?

$($a$)$ Compute the gradient of the loss $($b$)$ Randomly select a mini$-$batch $($X$,$Y$)$ from the training data $($c$)$ Initialize W $($d$)$ Update values of W $($e$)$ Using current values of W$,$ produce predictions Y$\_$hat

Choice $1$ of $4$:b$,$ a$,$ c$,$ d$,$ eChoice $2$ of $4$:c$,$ a$,$ b$,$ d$,$ eChoice $3$ of $4$:c$,$ b$,$ e$,$ a$,$ dChoice $4$ of $4$:b$,$ c$,$ e$,$ d$,$ a