Data Fundamentals (H) - Week 07 Quiz
1. Simulated annealing uses what metaheuristic to help avoid getting trapped in local minima?
Hill climbing.
Crossover rules.
Randomised restart.
A temperature schedule.
A population of solutions.
2. A
hyperparameter
of an optimisation algorithm is:
A measure of how good a solution is.
A direction in hyperspace.
The determinant of the Hessian.
A value that is used to impose constraints on the solution.
A value that affects how a solution is searched for.
3. First-order optimisation requires that objective functions be:
\(C^1\) continuous
one-dimensional
disconcerting
invertible
monotonic
4. The gradient vector \(\nabla L(\theta)\) is a vector which, at any given point \(\theta\) will:
have \(L_2\) norm 1
point in the direction of steepest descent
be zero
be equal to \(\theta\)
point towards the global minimum of \(L(\theta)\)
5. Finite differences is not an effective approach to apply first-order optimisation because:
the curse of dimensionality
of numerical roundoff issues.
all of the above
the effect of measurement noise
none of the above
6. Ant colony optimisation applies which two metaheuristics to improve random local search?
thants
random restart and hyperdynamics
gradient descent and crossover
temperature and memory
memory and population
Submit Quiz