Databricks ML Associate AutoML Workflows Quiz

1. What is the purpose of model interpretability tools in Databricks AutoML?

To increase model accuracy

To understand feature importance and model decisions

To eliminate the need for validation

To automatically select the best algorithm

2. How does Databricks AutoML determine the best model for a given dataset?

By training a single model and checking its accuracy

By training multiple models and comparing performance metrics

By randomly selecting an algorithm

By using only the oldest established algorithms

3. In Databricks AutoML, what does the model registry enable?

Version control and deployment of trained models

Automatic data cleaning

SQL query optimization

Real-time data streaming only

4. What is feature scaling and why is it important in AutoML?

Removing unnecessary features from the dataset

Normalizing or standardizing feature values for fair algorithm comparison

Increasing the dimensionality of the feature space

Converting continuous features to categorical

5. Which technique is used in AutoML to prevent model overfitting?

Increasing the number of features

Regularization, early stopping, and cross-validation

Removing all validation data

Using only the training dataset

6. In Databricks, MLflow Tracking is primarily used to:

Store raw datasets

Log parameters, metrics, and artifacts from ML experiments

Execute SQL queries

Manage cloud infrastructure

7. What is the role of train-test splitting in AutoML workflows?

To increase the total amount of training data

To evaluate model performance on unseen data

To reduce computational requirements

To eliminate feature engineering

8. How does Databricks AutoML typically handle class imbalance in classification tasks?

Ignores the imbalance during training

Applies techniques like oversampling, undersampling, or weighted losses

Removes the minority class entirely

Converts to regression instead

9. In AutoML workflows, what does model selection involve?

Choosing only the most complex model available

Evaluating multiple algorithms and selecting the best performer

Manually coding all machine learning algorithms

Using the first algorithm that trains successfully

10. Which of the following is a key advantage of using Databricks for AutoML?

It requires no data preparation

It integrates with Apache Spark for distributed computing

It works only with small datasets

It eliminates the need for data validation

11. What is the primary purpose of AutoML in Databricks?

To automatically generate code without human input

To automate feature engineering, model selection, and hyperparameter tuning

To replace data scientists entirely

To eliminate the need for data preprocessing

12. In Databricks AutoML, how are categorical variables typically handled?

Dropped from the dataset

Encoded using techniques like one-hot encoding or label encoding

Converted to numerical values without transformation

Kept as strings during model training

13. What role does Delta Lake play in Databricks ML workflows?

Provides ACID transactions and reliable data storage

Automatically trains machine learning models

Replaces MLflow for experiment tracking

Handles only unstructured data

14. Which metric is typically used to evaluate classification models in Databricks AutoML?

Mean Absolute Error

Accuracy, Precision, Recall, and F1-score

Root Mean Squared Error

Coefficient of determination

15. In AutoML workflows, cross-validation is used primarily to:

Split data into training and test sets

Assess model generalization and prevent overfitting

Increase the size of the training dataset

Normalize feature distributions

16. How does Databricks AutoML handle missing values in datasets?

Automatically removes rows with missing data

Uses imputation strategies like mean, median, or forward fill

Converts all missing values to zero

Ignores missing values during model training

17. What is hyperparameter tuning in the context of Databricks AutoML?

Manually adjusting model parameters during training

Automatically searching for optimal hyperparameter values

Removing underperforming features

Scaling features to a standard range

18. Which algorithm is commonly used by Databricks AutoML for regression tasks?

Random Forest and Gradient Boosting

K-means clustering

Principal Component Analysis

Apriori association rules

19. In Databricks AutoML, what does feature engineering typically include?

Creating new features from raw data and selecting relevant ones

Writing SQL queries only

Removing all categorical variables

Increasing the number of features indefinitely

20. Which Databricks feature enables tracking of machine learning experiments and models?

MLflow

Spark SQL

Delta Lake

Databricks Runtime