The core of ML training is gradient descent. MBZUAI tests if you truly understand derivatives in multiple dimensions.
The university offers a "Math for AI" refresher. The quizzes inside that bootcamp are the actual difficulty floor. If you can score 90% on those, you will pass.
Short answer / reasoning
This section tests your grasp of fundamental ML theory and model evaluation. Learning Paradigms
Let ( f(w) = \frac12 | y - Xw |^2 ), where ( y \in \mathbbR^n ), ( X \in \mathbbR^n \times d ), ( w \in \mathbbR^d ).
The sum of eigenvalues = Trace(A). Trace(A) = sum of diagonal entries.