Question 9
Domain 2: Describe fundamental principles of machine learning on AzureIn a machine learning context, what is the purpose of hyperparameter tuning?
Correct answer: A
Explanation
Hyperparameter tuning means selecting values for settings chosen before training, such as learning rate or tree depth, to improve model performance. It aims to "optimize the model's performance" by adjusting these settings so the model generalizes better to new data.
Why each option is right or wrong
A. To optimize the model's performance by adjusting its parameters
Hyperparameter tuning is the pre-training selection of model settings such as learning rate, regularization strength, number of trees, or maximum depth, and it is used to improve validation performance rather than to fit the training data directly. In practice, methods like grid search, random search, or Bayesian optimization compare candidate settings against a validation metric so the final model generalizes better to unseen data.
B. To preprocess and clean the data
C. To split data into training and testing sets
D. To assess the model's performance on different subsets of data