what can you plot to visualize if your GD algorithm is working?
plot a J (?) against #iterations. The group should slop downwards as interations increase
for a plot a J (?) against #iterations, how can you tell if GD has converged.
you can declare convergence of cost function J(?) if the next iteration or range of iterations decreases by less than a small value
for a plot a J (?) against #iterations, what should you do if the graph slopes up?
use smaller alpha. A sufficiently small ? step will ensure J(?) cost function decreases.
what is the learning rate?
the step alpha in gradient descent.

what happens if the learning rate alpha is too small?
slow convergence
what happens if the learning rate alpha is too big,
may not converge to the optima/minimum
how would you choose alpha?
try a range eg. .001, .003, .

01, .03..etc. going up 3 fold