$4.$ Choose all the valid descriptions about the stochastic gradient descent algorithm from the options below:

A$.$ Stochastic gradient descent $($SGD$)$ always finds the global optimal solution.

B$.$ Stochastic gradient descent $($SGD$)$ is typically more computationally efficient than standard gradient descent $($GD$).$

C$.$ Stochastic gradient descent $($SGD$)$ is typically more memory$-$demanding than stan$-$ dard gradient descent $($GD$).$

D$.$ Stochastic gradient descent $($SGD$)$ can be used to avoid bad local optima.

E$.$ Stochastic gradient descent $($SGD$)$ uses a fixed learning rate.