You decide to train your model using gradient descent. However, before you can turn your assignment in$,$ your

friends stop by to give you some suggestions.

$($a$)$ Lennie sees your regressor implementation and says there$$s a much simpler way! Just leave out sigmoids,

and let y $=$ w

T x $+$ b$.$ The derivatives are a lot less hassle and it runs faster. What$$s wrong with Lennie$$s

approach?

$($b$)$ George comes in and says that your logistic regressor is too complicated, but for a different reason. The

sigmoids are confusing and basically the same answer would be computed if we used sign functions, on the

right figure above, instead of the sigmoid. What$$s wrong with George$$s approach?