Gaussian Processes Quiz

Answers and explanations appear as you take the quiz.

1. A Gaussian Process defines a probability distribution over what?
   • Functions
   • Individual data points only
   • Discrete class labels
   • Neural network weights
2. Beyond a point estimate, what does a Gaussian Process return under its chosen kernel and noise assumptions?
   • Only an exact parameter vector
   • A posterior predictive distribution with an uncertainty estimate
   • A guaranteed calibrated interval under any model mismatch
   • A deterministic label with no variance
3. What role does the kernel (covariance function) play in a GP?
   • It labels the training data
   • It removes the need for training data
   • It encodes how similar outputs are for nearby inputs
   • It sets the learning rate
4. Why do standard Gaussian Processes struggle with very large datasets?
   • They cannot represent nonlinear functions
   • They require labeled classes
   • They ignore the training data entirely
   • Inverting the kernel matrix costs about n-cubed time
5. How are GP hyperparameters such as the length scale usually chosen?
   • By maximizing the marginal likelihood
   • By random guessing
   • By minimizing the number of data points
   • They are always fixed and never tuned
6. In which popular technique are Gaussian Processes a core component?
   • Convolutional image classification
   • Bayesian optimization for hyperparameter tuning
   • Tokenization of text
   • Backpropagation through deep networks
7. Which kernel is described in the guide as smoother than the Matérn alternative?
   • A unigram frequency kernel
   • A transition-probability kernel
   • The RBF (squared exponential) kernel
   • The Viterbi kernel
8. Why can a GP’s predictive uncertainty widen far from training observations?
   • The model adds more observed labels automatically
   • The mean function becomes a neural network
   • The kernel matrix becomes smaller than zero
   • There is less nearby data constraining the function under the selected covariance assumptions