Databricks-Machine-Learning-Professional 試験問題を無料オンラインアクセス

A Machine Learning Engineer is building a fraud detection model that needs to use both pre- computed features from a feature table and real-time calculated features based on user location data sent with each inference request. The engineer has created a Python UDF called calculate_distance in Unity Catalog at main.fraud_detection.calculate_distance that computes the distance between a transaction location and the user's current location. The feature table main.fraud_detection.user_features contains historical user spending patterns with primary key user_id.
The engineer has written the following code to implement this scenario:

Which benefit of this implementation approach makes it suited to the real-time fraud detection use case?

• A.The model automatically performs feature lookup and computation during inference without additional serving code.
• B.The FeatureLookup function avoids the need for joining the full table main.fraud_detection.user_features with training_set increasing efficiency.
• C.The FeatureFunction caches computed distance values in the online store to improve inference latency for location pairs.
• D.The Unity Catalog registry automatically creates REST API endpoints for UDF functions used in feature computation.

A Machine Learning Engineer needs to develop fraud detection models with Databricks. They need to ensure seamless collaboration between data engineers and data scientists while maintaining strict governance, version control, and traceability as models progress from development to production. So, they need to choose the Databricks feature that will enable centralized model lineage tracking, cross-workspace access control, and automated synchronization of model versions with their training data. Which Databricks feature will do this?

• A.Mosaic AI Model Serving for deploying models as REST APIs
• B.Lakehouse Monitoring for tracking data quality and model drift
• C.Unity Catalog for governing models via a centralized registry with role-based access
• D.MLflow Tracking for logging parameters and metrics during experimentation

A Machine Learning Engineer needs to deploy a custom model using Databricks Model Serving.
The model requires an external tokenizer file (for example, a vocabulary or pre-trained tokenizer) to function correctly. They need to ensure this tokenizer file is included with the model so it is available during model serving. How should they package this tokenizer file as part of the model deployment?

• A.Include tokenizer file as part of the model's software dependencies by adding it to the conda_env specification.
• B.Access the tokenizer file directly from Volumes at model inference time, which requires pre- provisioning access to the Volume for every model serving endpoint.
• C.Use the artifacts parameter in mlflow.pyfunc.log_model() to package the tokenizer file, referencing its path in Volumes.
• D.Use the code_paths parameter in mlflow.pyfunc.log_model() to include the tokenizer file, assuming it's part of the custom Python logic needed by the model.

A Machine Learning Engineer has trained a credit scoring model and needs to evaluate fairness metrics across different customer segments while maintaining different levels of granularity for business reporting. They need to compute metrics like precision, recall, and demographic parity at the individual feature level (credit_score_range, income_bracket) as well as intersectional slices (combinations of features). The model outputs are stored in a Delta table with prediction probabilities and actual default labels. The engineer wants to systematically evaluate model performance across these various feature slices and granularities. Which approach will do this?

• A.Create Lakehouse Monitoring with slicing expressions for individual features and intersection conditions.
• B.Use Unity Catalog metric views with dimensions defined for each feature and measures for the fairness metrics.
• C.Build separate Spark DataFrames for each slice and compute metrics using standard DataFrame operations.
• D.Implement custom MLflow evaluation functions that iterate through all possible feature combinations.

A Machine Learning Engineer is tasked with implementing a simple solution, requiring as little code as possible, to monitor a regression model and ensure that the R2 score does not exceed a certain threshold. If the threshold is exceeded, the solution needs to send an email to a distribution list. Which proposed solution meets the criteria?

• A.Configure Lakehouse Monitoring using the Inference profile type. Create a SQL Alert that runs on a schedule and evaluates if the R2 score in the profile table exceeds the threshold. If so, the Alert should send an email to the configured email notification destination.
• B.Configure Lakehouse Monitoring using the Inference profile type. Create a SQL Query that runs on a schedule and evaluates if the R2 score in the profile table exceeds the threshold. If so, the Alert should send an email to the configured email notification destination.
• C.Create a Workflow that runs on a schedule. The Workflow executes a SQL Query that calculates the R2 score. Use Workflow If/Else condition to execute a Notebook Task that sends the email if the R2 score threshold is exceeded.
• D.Configure Lakehouse Monitoring using the Inference profile type. Create a Workflow that runs on a schedule and executes a SQL query that evaluates the R2 score in the profile table. Use Workflow If/Else condition to execute a Notebook Task that sends the email if the R2 score threshold is exceeded.