Neural Nuisance

For Harry, multi-choice questions were always challenging.

He was much better with words and always found a way to reason while elaborating on his answers. But multi-choice questions were cold and to the point. They were an all-or-nothing affair.

Harry doesn't have a choice. If he wants to graduate, he has to answer the question:

Which of the following sentences are true about neural networks?

Neural networks can find the optimal solution for convex problems.

Neural networks can find the optimal solution for concave problems.

Neural networks can approximate any function when using non-linear activations.

We can only optimize neural networks using the Gradient Descent algorithm.