A security team wants to use Snowflake ML functions to detect unusual login activity. They have a table of login events with columns for user, timestamp, IP address, and success flag. They want to flag logins that deviate from a user's typical pattern. Which approach should they use?

Run SNOWFLAKE.ML.ANOMALY_DETECTION on the login events table, partitioned by user, with the timestamp column as the time series.

Use a SQL query with a simple threshold on the number of logins per hour to flag unusual activity.

Run SNOWFLAKE.ML.FORECAST on the login events table to predict the next login time for each user.

Create a classification model using SNOWFLAKE.ML.CLASSIFICATION to label each login as normal or anomalous.

A security team is using SNOWFLAKE.ML.ANOMALY_DETECTION to monitor access to a sensitive database. The data is a time series of the number of queries per hour. The team notices that the model is flagging too many false positives during business hours. They suspect the model is not accounting for the regular daily pattern of high activity. What should they do to improve the model's accuracy?

Increase the 'sensitivity' parameter of the anomaly detection function to make it less sensitive.

Aggregate the data into daily totals instead of hourly.

Use SNOWFLAKE.ML.FORECAST instead to predict the expected number of queries and compare it to the actual.

Remove the timestamp column and use only the query count as the input.

Which Snowflake feature allows you to track who is using AI/ML functions and what queries they are running?

Query history and access history

A data engineer is preparing a dataset for training a classification model to identify malicious files. The dataset contains a column 'file_size' with some NULL values and a column 'file_type' with values like 'exe', 'pdf', and 'docx'. What is the best way to prepare this data for the model?

Use a SQL query to cast 'file_type' to a numeric value and leave 'file_size' as is.

Remove all rows with NULL values in 'file_size' and keep 'file_type' as a string.

Impute the NULL values in 'file_size' with the mean and use one-hot encoding for 'file_type'.

Replace NULL values in 'file_size' with 0 and use a label encoder for 'file_type'.

What is a best practice for securing AI/ML models deployed in Snowflake?

Applying role-based access controls to limit who can use the model

Storing model data in plain text for simplicity

Disabling all auditing on model usage

Granting all users access to the model for transparency

Snowflake SnowPro Advanced:Security Engineer

Domain 5Objective 1

Leverage Snowflake AI/ML Capabilities to Enhance Security Posture SNOWPRO-ADVANCED-SECURITY-ENGINEER Practice Questions (Page 1)

Part of the AI/ML Security domain, which makes up ~9% of our current practice bank.

27questions here

6free pages

8concepts

Questions 1–5

1application · medium
A security team wants to use Snowflake ML functions to detect unusual login activity. They have a table of login events with columns for user, timestamp, IP address, and success flag. They want to flag logins that deviate from a user's typical pattern. Which approach should they use?
Select an answer first
2expert · hard
A security team is using SNOWFLAKE.ML.ANOMALY_DETECTION to monitor access to a sensitive database. The data is a time series of the number of queries per hour. The team notices that the model is flagging too many false positives during business hours. They suspect the model is not accounting for the regular daily pattern of high activity. What should they do to improve the model's accuracy?
Select an answer first
3foundation · easy
Which Snowflake feature allows you to track who is using AI/ML functions and what queries they are running?
Select an answer first
4application · medium
A data engineer is preparing a dataset for training a classification model to identify malicious files. The dataset contains a column 'file_size' with some NULL values and a column 'file_type' with values like 'exe', 'pdf', and 'docx'. What is the best way to prepare this data for the model?
Select an answer first
5foundation · easy
What is a best practice for securing AI/ML models deployed in Snowflake?
Select an answer first

Finished these 5 questions?

Review the revealed explanations, or continue through the curriculum.

Next 5 questions →Back to SNOWPRO-ADVANCED-SECURITY-ENGINEER

Free Basic Practice is a study aid with revealable answers — not a scored exam. Examers.io is independent and not affiliated with or endorsed by Snowflake. “SNOWPRO-ADVANCED-SECURITY-ENGINEER” is a trademark of its owner, used for identification only.