You've built a customer churn prediction model in Snowflake, and are using the AUC as your primary performance metric. You notice that your model consistently performs well (AUC > 0.85) on your validation set but significantly worse (AUC < 0.7) in production. What are the possible reasons for this discrepancy? (Select all that apply)

• A. Your training and validation sets are not representative of the real-world production data due to sampling bias.
• B. The AUC metric is inherently unreliable and should not be used for model evaluation.
• C. The production environment has significantly more missing data compared to the training and validation environments.
• D. Your model is overfitting to the validation data. This causes to give high performance on validation set but less accurate in the real world.
• E. There's a temporal bias: the customer behavior patterns have changed since the training data was collected.

You are managing a machine learning model lifecycle in Snowflake using the Model Registry. Which of the following statements are true regarding model lineage and governance when utilizing the Model Registry for model versioning and deployment?

• A. The Model Registry automatically tracks the exact SQL queries used to train the model, allowing for full reproducibility of the training process.
• B. Custom tags and metadata can be associated with each model version, enabling detailed documentation and traceability of model development and deployment.
• C. Model Registry automatically retrains models based on scheduled data updates, ensuring models are always up-to-date without manual intervention.
• D. The Model Registry provides a central repository to register, version, and manage models, enabling better collaboration and governance across data science teams.
• E. Integration with Snowflake's RBAC (Role-Based Access Control) allows for granular control over who can register, update, and deploy model versions.

A marketing team at 'RetailSphere' wants to segment their customer base using unstructured textual data (customer reviews) stored in a Snowflake VARIANT column named 'REVIEW TEXT within the table 'CUSTOMER REVIEWS'. They aim to identify distinct customer segments based on sentiment and topics discussed in their reviews. They want to use a Supervised Learning approach for this task. Which of the following strategies best describes the appropriate approach within Snowflake, considering performance and scalability? Assume you have pre-trained sentiment and topic models deployed as Snowflake external functions.

• A. Create a Snowflake external function to call a pre-trained sentiment analysis and topic modeling model hosted on AWS SageMaker. Apply these functions to the ' REVIEW_TEXT column to generate sentiment scores and topic probabilities. Subsequently, use these features as input to a supervised classification model (e.g., XGBoost) also deployed as a Snowflake external function, training on a manually labeled subset of reviews.
• B. Extract the ' REVIEW_TEXT column, manually categorize a small subset of reviews into predefined segments. Train a text classification model (e.g., using scikit-learn) externally, deploy it as a Snowflake external function, and then apply this function to the entire 'REVIEW TEXT column to predict segment assignments. Manually adjust cluster centroids to represent the manually labeled dataset.
• C. Create a Snowflake external function to call a pre-trained sentiment analysis and topic modeling model hosted on Azure ML. Apply these functions to the REVIEW_TEXT column to generate sentiment scores and topic probabilities. Subsequently, use these features as input to an unsupervised clustering algorithm (e.g., DBSCAN) within Snowflake, relying solely on data density to define segments.
• D. Extract the 'REVIEW TEXT column, apply sentiment analysis and topic modeling using Java within a Snowflake UDF, and then perform hierarchical clustering directly on the resulting features within Snowflake. Manually label the clusters after visual inspection.
• E. Extract the column, apply sentiment analysis and topic modeling using Python within a Snowflake UDF, and then perform K-Means clustering directly on the resulting features within Snowflake. Define the labels after clustering based on the majority class of the topics and sentiments in each cluster.

A retail company is using Snowflake to store transaction data'. They want to create a derived feature called 'customer _ recency' to represent the number of days since a customer's last purchase. The transactions table 'TRANSACTIONS has columns 'customer_id' (INT) and 'transaction_date' (DATE). Which of the following SQL queries is the MOST efficient and scalable way to derive this feature as a materialized view in Snowflake?

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