Azure Data Scientist AutoML Configuration

1. When configuring AutoML for imbalanced classification data, which approach helps improve results?

Setting class_weight to 'balanced' and using appropriate metrics like AUC-weighted

Removing the minority class entirely

Using only linear models

Increasing the training data size without addressing imbalance

2. When configuring AutoML for a multi-class classification problem, which primary_metric balances precision and recall across all classes?

Weighted F1-score or Macro F1-score

Binary accuracy

Mean Squared Error

Specificity

3. In Azure AutoML, what is the purpose of the 'train_test_ratio' or validation_size parameter?

To define the proportion of data used for model training versus validation

To set the number of algorithms to test

To control the model's complexity level

To specify the maximum experiment duration

4. Which of the following is a valid approach for handling missing values in AutoML preprocessing?

AutoML can automatically detect and impute missing values using mean, median, or forward-fill methods

Missing values must be manually removed before feeding data to AutoML

AutoML ignores missing values entirely

Missing values automatically become zero in all cases

5. In AutoML configuration, what is the benefit of setting 'experiment_exit_score'?

To stop the experiment once a model achieves a specified performance threshold

To force the experiment to continue indefinitely

To remove poorly performing models from consideration

To increase the experiment's total duration

6. What does the 'max_concurrent_iterations' parameter control in AutoML?

The maximum number of model training iterations running in parallel

The total number of models to train sequentially

The number of features to engineer

The size of each training batch

7. Which metric is typically used as the primary_metric for regression tasks in AutoML?

Root Mean Squared Error (RMSE) or R-squared

Precision and Recall

Confusion Matrix

F1-score

8. What is the purpose of setting 'featurization' to 'auto' in AutoML configuration?

To allow AutoML to automatically apply feature engineering transformations

To disable all feature preprocessing

To manually define all features

To remove outliers from the dataset

9. In time series forecasting with AutoML, what is the significance of the 'forecast_horizon' parameter?

It specifies the number of time periods ahead that the model should predict

It sets the training data time range

It controls the frequency of data collection

It determines the model's learning rate

10. What does the 'enable_early_stopping' parameter do in AutoML experiments?

Terminates the experiment early if no improvement is observed in model performance

Prevents the experiment from running at all

Forces the experiment to run for maximum duration

Removes all preprocessing steps

11. What is the primary purpose of Azure AutoML in the context of machine learning model development?

To automate the entire ML pipeline from data ingestion to model deployment

To replace data scientists with automated processes

To manually select only linear regression models

To eliminate the need for data preprocessing

12. What is the role of the 'max_cores_per_iteration' parameter in AutoML configuration?

To limit CPU resources used by each individual model training iteration

To set the total experiment duration

To specify the number of models to train

To control the size of the validation set

13. In Azure AutoML, what does the 'n_cross_validations' parameter control?

The number of folds used in cross-validation

The number of training epochs

The number of algorithms to test

The number of data preprocessing steps

14. Which feature engineering technique can AutoML apply to improve model performance?

Automatic feature scaling and normalization

Manual selection of all features by the data scientist

Removal of all numeric features

Converting all features to categorical

15. What is the purpose of setting 'blocked_models' in an AutoML configuration?

To exclude specific algorithms from being considered during training

To increase the training speed

To improve model accuracy automatically

To reduce memory consumption

16. In AutoML classification tasks, what does the 'primary_metric' parameter determine?

The metric used to evaluate and rank models during the experiment

The minimum accuracy required to stop the experiment

The number of features to use in the model

The size of the training dataset

17. Which validation type in AutoML divides data into k subsets for iterative testing?

Train-test split

K-fold cross-validation

Stratified sampling

Time series split

18. What is cross-validation used for in AutoML experiments?

To assess model performance on unseen data and reduce overfitting risk

To increase the training data size

To replace the need for a separate test set

To automatically select the best algorithm

19. In Azure AutoML, what does the 'task' parameter specify?

The type of machine learning problem (classification, regression, forecasting)

The number of CPU cores to allocate

The specific algorithm to use for training

The percentage of data reserved for testing

20. Which parameter in AutoML experiment configuration controls the maximum amount of time the experiment runs?

Iteration_timeout_minutes

Experiment_timeout_minutes

Total_training_time

Max_duration_seconds

Azure Data Scientist AutoML Experiment Configuration Quiz