Explain the logic behind this loss function. In particular, why do we use the $$log$($y$)$

here?

$($k$)($Writing$)[10$ pt$]$ Finally, it is time for you to improve your intuition about how hyperparameters

affect the training and testing of NNs$.$ In this section, vary hyperparameters to achieve your highest possible

test set performance. Report what you changed, to which values, and your final test performance value.

$4$