Leverage Snowflake AI/ML Capabilities to Enhance Security Posture SNOWPRO-ADVANCED-SECURITY-ENGINEER Practice Questions (Page 5)

Questions 21–25

1. 21foundation · easy

   What is the typical output of a classification model when applied to a security event?

   • AA list of contributing factors
   • BA forecasted future value
   • CA predicted class label for the event
   • DAn anomaly score
   Reveal Solution Discussion Select an answer first
2. 22expert · hard

   A security team wants to use an external AI/ML platform to analyze user behavior data stored in Snowflake. The platform needs to receive a daily export of the data. The team is concerned about data security and wants to ensure that only the necessary data is exported and that the export process is auditable. Which approach should they use?

   • ACreate a secure view that contains only the necessary columns and use a task to export the view's data to a stage, then use an external function to send it to the platform.
   • BCreate a materialized view of the data and grant the external platform SELECT on it.
   • CUse Snowpipe to continuously load the data into the external platform's storage.
   • DGrant the external platform direct access to the underlying tables using a data share.
   Reveal Solution Discussion Select an answer first
4. 23foundation · easy

   What is a key step in preparing security data for AI/ML model training in Snowflake?

   • ADeleting all historical data
   • BEncrypting all data with a single key
   • CStoring data in unstructured format
   • DCleaning and transforming the data into a suitable format
   Reveal Solution Discussion Select an answer first
5. 24application · medium

   A security administrator needs to ensure that the use of Snowflake Cortex functions by data scientists is compliant with internal policies. They want to know which users are calling Cortex functions, how often, and what data is being processed. Which Snowflake feature should they use?

   • AUse the INFORMATION_SCHEMA.TABLES view to see which tables are being accessed.
   • BEnable Snowflake's data masking policy on the Cortex functions.
   • CQuery the ACCOUNT_USAGE.QUERY_HISTORY view to see all queries, including those that call Cortex functions.
   • DCreate a task that runs every hour to list all active sessions.
   Reveal Solution Discussion Select an answer first
6. 25application · medium

   A security operations team needs to automatically categorize incoming security alerts as 'malicious', 'suspicious', or 'benign'. They have a historical dataset of alerts with labels assigned by analysts. They want to build and deploy a model in Snowflake that can classify new alerts as they arrive. Which approach should they take?

   • AUse SNOWFLAKE.ML.FORECAST to predict the number of malicious alerts expected in the next hour.
   • BUse SNOWFLAKE.ML.CLASSIFICATION to train a model on the labeled historical data and use the resulting model to classify new alerts.
   • CUse SNOWFLAKE.ML.ANOMALY_DETECTION to identify alerts that are different from the historical baseline.
   • DExport the labeled data to an external tool, train a model there, and import the model back into Snowflake for inference.
   Reveal Solution Discussion Select an answer first