Snowflake DSA-C03 Exam : SnowPro Advanced: Data Scientist Certification Exam

You've trained a model using Snowflake ML and want to deploy it for real-time predictions using a Snowflake UDF. To ensure minimal latency, you need to optimize the UDF's performance. Which of the following strategies and considerations are most important when creating and deploying a UDF for model inference in Snowflake to minimize latency, especially when the model is large (e.g., > 100MB)?
Select all that apply.

• A. Store the trained model as a BLOB within the UDF code itself to avoid external dependencies.
• B. Ensure the UDF code is written in Python and utilizes vectorized operations with libraries like NumPy to process data in batches efficiently.
• C. Utilize a Snowflake external function instead of a UDF if the model requires access to resources outside of Snowflake's environment.
• D. Use smaller warehouse size for UDF evaluation in order to reduce latency and compute costs.
• E. Use a Snowflake Stage to store the model file and load the model within the UDF using 'snowflake.snowpark.files.SnowflakeFile' to minimize memory footprint.

You are working with a dataset containing customer reviews for various products. The dataset includes a 'REVIEW TEXT column with the raw review text and a 'PRODUCT ID' column. You want to perform sentiment analysis on the reviews and create a new feature called 'SENTIMENT SCORE for each product. You plan to use a UDF to perform the sentiment analysis. Which of the following steps and SQL code snippets are essential for implementing this feature engineering task in Snowflake, ensuring optimal performance and scalability? Select all that apply:

• A. Apply the sentiment analysis UDF to the 'REVIEW TEXT column within a 'SELECT statement, grouping by 'PRODUCT ID and calculating the average 'SENTIMENT_SCORE' using
• B. Ensure the UDF is vectorized to process batches of reviews at once, improving performance. This can be achieved using decorator on top of the python function.
• C. Use the 'SNOWFLAKE.ML' package to train a sentiment analysis model directly within Snowflake, eliminating the need for a separate UDF.
• D. Cache the results of the sentiment analysis UDF in a temporary table to avoid recomputing the scores for the same reviews in subsequent queries. Use 'CREATE TEMPORARY TABLE to create a temporary table.
• E. Create a Python UDF that takes the 'REVIEW_TEXT as input and returns a sentiment score (e.g., between -1 and 1). Then, use 'CREATE OR REPLACE FUNCTION' statement to register the UDF.

You are building a customer churn prediction model for a telecommunications company. You have a 'CUSTOMER DATA' table with a 'MONTHLY SPENDING' column that represents the customer's monthly bill amount. You want to binarize this column to create a feature indicating whether a customer is a 'High Spender' or 'Low Spender'. You decide that customers spending more than $75 are 'High Spenders'. Which of the following Snowflake SQL statements is the most efficient and correct way to achieve this, considering performance and readability, while avoiding potential NULL values in the resulting binarized column?

• A. Option A
• B. Option D
• C. Option B
• D. Option E
• E. Option C

You are building a fraud detection model using Snowflake data'. One of the features is 'transaction_amount', which has a highly skewed distribution and contains outlier values. Which scaling technique is most appropriate to handle this situation effectively in Snowflake, considering the need to minimize the impact of outliers and preserve the shape of the distribution as much as possible, before feeding the data into a machine learning model? Assume you have sufficient compute resources.

• A. RobustScaler (using interquartile range)
• B. Power Transformer (Yeo-Johnson or Box-Cox)
• C. No scaling is needed as tree-based models are robust to skewed data.
• D. MinMaxScaler (Min-Max scaling)
• E. StandardScaler (Z-score normalization)

You've deployed a fraud detection model in Snowflake. The model is implemented as a Python UDF that uses a pre-trained scikit-learn model stored as a stage file. Your goal is to enable near real-time fraud detection on incoming transactions. Due to regulatory requirements, you need to maintain a detailed audit trail of all predictions, including the input features, model version, prediction scores, and any errors encountered during the prediction process. Which of the following approaches are valid and efficient for storing these audit logs and predictions in Snowflake?

• A. Use Snowflake's 'SYSTEM$QUERY LOG' table to extract information about the UDF execution and join it with the transaction data to reconstruct the audit trail.
• B. Create a dedicated table with columns for transaction ID, input features (as a JSON VARIANT), model version, prediction score, error message (if any), and prediction timestamp. Use a Snowflake Sequence to generate unique log IDs.
• C. Store the audit logs as unstructured text files in an external stage (e.g., AWS S3) and periodically load them into a Snowflake table using COPY INTO command.
• D. Utilize Snowflake's Streams and Tasks to automatically capture changes to the transaction table and trigger the prediction UDF, storing the audit logs in a separate table with similar structure as described in option A.
• E. Log the audit information to an external logging service (e.g., Splunk) using an external function called from within the UDF.