Tune Model Hyperparameters SNOWPRO-ADVANCED-DATA-SCIENTIST Practice Questions (Page 5)

Questions 21–25

1. 21application · easy

   A data scientist is explaining to a colleague the difference between model parameters and hyperparameters. Which statement correctly describes this distinction?

   • ABoth are learned from the data during training.
   • BBoth are set before training and remain fixed.
   • CHyperparameters are learned from the data, while model parameters are set before training.
   • DModel parameters are learned from the data, while hyperparameters are set before training.
   Reveal Solution Discussion Select an answer first
2. 22expert · hard

   A data scientist is using Bayesian optimization to tune hyperparameters for a model. They notice that the optimization is spending too many trials exploring unpromising regions. What adjustment could improve the efficiency of the search?

   • AIncrease the number of initial random trials
   • BUse a random search instead
   • CIncrease the exploration parameter (e.g., kappa in Expected Improvement)
   • DDecrease the exploration parameter (e.g., kappa in Expected Improvement)
   Reveal Solution Discussion Select an answer first
4. 23foundation · easy

   If a model shows poor performance on both the training and validation sets, which adjustment to the hyperparameter search space is most likely to help?

   • ADecrease the maximum depth of trees
   • BReduce the number of training epochs
   • CIncrease the model's capacity (e.g., deeper trees, more layers)
   • DIncrease the regularization strength
   Reveal Solution Discussion Select an answer first
5. 24foundation · easy

   Which of the following is a key advantage of Bayesian optimization over grid search?

   • AIt is guaranteed to find the best possible hyperparameters
   • BIt can find the global optimum in a single iteration
   • CIt requires no validation data
   • DIt typically needs fewer evaluations to find good hyperparameters
   Reveal Solution Discussion Select an answer first
6. 25foundation · easy

   Which of the following best distinguishes a hyperparameter from a model parameter?

   • AHyperparameters are fixed constants; model parameters vary with the random seed
   • BHyperparameters are set before training; model parameters are learned during training
   • CHyperparameters are learned from the training data; model parameters are set by the user
   • DHyperparameters only apply to neural networks; model parameters apply to all algorithms
   Reveal Solution Discussion Select an answer first