DSA-C03 試験問題を無料オンラインアクセス

You are evaluating a binary classification model built in Snowflake for predicting customer churn. You have access to the model's predictions on a holdout dataset, and you want to use both the ROC curve and the confusion matrix to comprehensively assess its performance. Which of the following statements regarding the interpretation and use of ROC curves and confusion matrices are correct in this scenario?

• A.While the ROC curve is independent of the class distribution, the metrics derived from the confusion matrix (e.g., precision, recall) can be significantly affected by imbalanced datasets.
• B.The confusion matrix allows you to calculate precision, recall, F I-score, and accuracy, which are all useful for understanding the model's performance in terms of correctly and incorrectly classified instances.
• C.The area under the ROC curve (AUC) provides a single scalar value representing the overall discriminatory power of the model, with a higher AUC indicating better performance. An AUC of 0.5 indicates that the model performs no better than random chance.
• D.In Snowflake, you can generate ROC curves and confusion matrices directly using the 'SYSTEM$PREDICT function with appropriate parameters and visualizing the results using a tool like Snowsight or Tableau.
• E.The ROC curve visualizes the trade-off between true positive rate (sensitivity) and false negative rate (1 - specificity) at various threshold settings.

You are working on a fraud detection model and need to prepare transaction data'. You have two tables: 'transactions' (transaction_id, customer_id, transaction_date, amount, merchant_id) and (merchant_id, city, state). You need to perform the following data cleaning and feature engineering steps using Snowpark: 1. Remove duplicate transactions based on 'transaction_id'. 2.
Join the 'transactions' table with the 'merchant_locations table to add city and state information to each transaction. 3. Create a new feature called 'amount_category' based on the transaction amount, categorized as 'Low', 'Medium', or 'High'. 4. The categorization thresholds are defined as follows: 'LoW: amount < 50 'Medium': 50 amount < 200 'High': amount >= 200 Which of the following statements about performing these operations using Snowpark are accurate?

• A.Removing duplicate transactions can be efficiently done using the method on the Snowpark DataFrame, specifying 'transaction_id' as the subset. Creating the amount categories can be completed using the 'when' clause with multiple 'otherwise' clauses.
• B.A LEFT JOIN should be used to join the 'transactions' and 'merchant_location' tables to ensure that all transactions are included, even if some merchant IDs are not present in the 'merchant_location' table.
• C.The construct in Snowpark can be used to create the 'amount_category' feature directly within the DataFrame transformation without needing a UDF
• D.Removing duplicate transactions can be efficiently done using the method on the Snowpark DataFrame, specifying 'transaction_id' as the subset. Creating the amount categories requires use of a User-Defined Function (UDF) as the logic can't be efficiently embedded in a single 'when' clause.
• E.You can register SQL UDF to calculate the 'amount_category' using 'CASE WHEN' statement

You are developing a Snowflake Native App that leverages Snowflake Cortex for text summarization. The app needs to process user-provided text input in real-time and return a summarized version. You want to expose this functionality as a secure and scalable REST API endpoint within the Snowflake environment. Which of the following strategies are MOST suitable for achieving this, considering best practices for security and performance?

• A.Utilize a Snowflake Stored Procedure written in SQL that invokes the 'SNOWFLAKE.CORTEX.SUMMARIZE' function, and then create a Snowflake API Integration to expose the stored procedure as a REST endpoint.
• B.Create a Snowflake External Function using Python that directly calls the 'SNOWFLAKE.CORTEX.SUMMARIZE' function and expose this function via a REST API gateway outside of Snowflake.
• C.Develop a Snowflake Native App that includes a Java UDF that calls 'SNOWFLAKE.CORTEX.SUMMARIZE and expose a REST API using Snowflake's built-in REST API capabilities within the Native App framework.
• D.Write a Snowflake Stored Procedure using Javascript to invoke the 'SNOWFLAKE.CORTEX.SUMMARIZE function, deploy the procedure to a Snowflake stage, and then trigger it via an AWS Lambda function integrated with Snowflake.
• E.Develop a Snowflake Native App containing a Python UDF that calls 'SNOWFLAKCORTEX.SUMMARIZE function, and expose it as a REST API endpoint using Snowflake's API Integration feature within the app package.

You have built a customer churn prediction model using Snowflake ML and deployed it as a Python stored procedure. The model outputs a churn probability for each customer. To assess the model's stability and potential business impact, you need to estimate confidence intervals for the average churn probability across different customer segments. Which of the following approaches is MOST appropriate for calculating these confidence intervals, considering the complexities of deploying and monitoring models within Snowflake?

• A.Use a separate SQL query to extract the churn probabilities and customer segment information from the table where the stored procedure writes its output. Then, use a statistical programming language like Python (outside of Snowflake) to calculate the confidence intervals for each segment.
• B.Implement a custom SQL function to approximate confidence intervals based on the Central Limit Theorem, assuming the churn probabilities are normally distributed.
• C.Pre-calculate confidence intervals during model training and store them as metadata alongside the model in Snowflake. This avoids runtime computation.
• D.Calculate a single confidence interval for the overall average churn probability across all customers. Customer segmentation confidence intervals are statistically invalid and not applicable for Snowflake ML models.
• E.Calculate confidence intervals directly within the Python stored procedure using bootstrapping techniques and appropriate libraries (e.g., scikit-learn) before returning the churn probability.

You have trained a linear regression model in Snowpark ML to predict house prices. After training, you want to assess the overall feature importance using the model's coefficients. Consider the following Snowflake table containing the coefficients:

Which of the following statements are correct interpretations of these coefficients regarding feature impact?

• A.An increase of one square foot (sqft) in house size is associated with an increase of $120.5 in the predicted house price.
• B.The 'age' feature has an insignificant impact because its coefficient is small.
• C.The 'location_score' feature is the most influential predictor in determining house price.
• D.The 'bedrooms' feature has a positive impact on the house price since the coefficient is negative.
• E.Increasing the number of bedrooms is associated with a decrease in the predicted house price.