Hyperparameter Tuning in Machine Learning Quiz

The primary goal of hyperparameter tuning in machine learning is ....

• To select the optimal features for
• To select the optimal hyperparameter
• To increase model complexity
• To choose the optimal training data

Which of the following is a method of hyperparameter optimization that exhaustively tries every possible combination of hyperparameters?

• Random Search

• Gradient Search

• Grid Search

• Bayesian Optimization

What is a major disadvantage of using Grid Search?

• It guarantees finding the best hyperparameters

• It can be very slow

• It requires manual intervention

• It only works with classification models

Which of the following is a key advantage of Random Search over Grid Search?

• It guarantees the best hyperparameters

• It tests all possible combinations of hyperparameters

• It is faster and can find good hyperparameters in fewer iterations

• It is more reliable in finding the global optimum

What does the scoring parameter specify in GridSearchCV or RandomizedSearchCV?

• The method of hyperparameter selection

• The evaluation metric used to compare models

• The number of iterations for grid search

• The splitting method for cross-validation

Why does Scikit-Learn use negative values for MSE (neg_mean_squared_error) in GridSearchCV?

• To minimize the value while fitting the model

• To maximize the value during optimization

• To make it easier to calculate R-squared

• To avoid overfitting

What is Stratified K-Fold Cross-Validation useful for?

• Ensuring that each fold has an equal distribution of the target variable in imbalanced datasets

• Training on all data points at once

• Preventing the model from overfitting

• Decreasing the number of training data

In Python, which library would you use to save a trained model? (Choose 2)

• Worklib

• Onion

• Joblib

• Pickle

Which of the following is a key advantage of using Random Search for hyperparameter tuning over Grid Search?

• Random Search is more accurate in finding the best parameters

• Random Search tests all possible combinations of hyperparameters

• Random Search is computationally faster and often finds good parameters more efficiently

• Random Search always guarantees better results than Grid Search