Study with DSA-C03 most valid questions & verified answers

Snowflake DSA-C03 Exam Overview:

Snowflake DSA-C03 Exam Syllabus Topics:

Snowflake SnowPro Advanced: Data Scientist Certification Sample Questions:

Question 1

You are building a machine learning model to predict customer churn for a telecommunications company. One of the features is 'tariff_plan', which is a string representing different tariff plans (e.g., 'Basic', 'Premium', 'Unlimited'). You need to encode this feature for your model, but you also want to handle potential new tariff plans that might appear in future data'. Which encoding method and Snowflake SQL approach would be MOST suitable to minimize dimensionality and address unseen values effectively, assuming the number of plans is moderately high (around 20-30)?

A. Label Encoding using a UDF (User-Defined Function) with a predefined mapping, assigning a new integer to unseen values, and storing the mapping in a separate table in Snowflake.
B. Hash Encoding (Feature Hashing) using a UDF in Snowflake, with a fixed number of features and a hashing function to map each tariff plan to a feature index, accepting potential collisions. Handle new tariff plans naturally through the hashing function.
C. Target Encoding (Mean Encoding) using Snowflake SQL, calculating the mean churn rate for each tariff plan and using that as the encoded value. Handle unseen values with the global mean churn rate, being mindful of potential target leakage.
D. Binary Encoding using a UDF to convert each tariff plan into binary code, storing encoded results into snowflake, then splitting the binary representation into separate columns.
E. One-Hot Encoding using CREATE OR REPLACE VIEW, handling new values by NULLIF('Unknown', tariff_plan) before encoding, potentially leading to a high number of columns.

Question 2

You are a data scientist working with a Snowflake table named 'CUSTOMER TRANSACTIONS' that contains sensitive PII data, including customer names and email addresses. You need to create a representative sample of 1% of the data for model development, ensuring that the sample is anonymized and protects customer privacy. The sample must be reproducible for future model iterations.
Which of the following steps are most appropriate using Snowpark for Python and SQL?

A. Use the 'QUALIFY OVER (ORDER BY RANDOM()) (SELECT COUNT( ) 0.01 FROM CUSTOMER_TRANSACTIONS)' clause with SHA256 on sensitive columns directly within a CREATE TABLE AS statement to generate an anonymized sample. The function should return only 1 percentage of row.
B. Create a new table using 'CREATE TABLE AS SELECT statement combined with 'SAMPLE clause and SHA256 hashing functions in SQL to create the sample and anonymize data. Manually seed the random number generator in Python before executing the SQL statement via Snowpark.
C. Use the 'SAMPLE clause in a SQL query to extract 1% of the rows, then apply SHA256 hashing to the 'customer_name' and 'email_addresS columns within Snowpark using a UDF. Seed the sampling for reproducibility.
D. Employ stratified sampling based on a customer segment column, then anonymize data. Use the TABLESAMPLE BERNOULLI function in SQL with a 1 percent sample rate. Apply SHA256 hashing to the 'customer_name' and 'email_addresS columns using SQL functions.
E. Use Snowpark DataFrame's 'sample' function with a fraction of 0.01 and a fixed random seed. Before sampling, create a view that masks 'customer_name' and 'email_address' columns, and then sample from the view.

Question 3

A data scientist needs to analyze website session data stored in a Snowflake table named 'WEB SESSIONS'. The table contains columns like 'SESSION D', 'USER_ID, 'PAGE_VIEWS', 'TIME SPENT_SECONDS', and 'TIMESTAMP. They want to identify potential bot traffic by analyzing the correlation between 'PAGE VIEWS' and 'TIME SPENT SECONDS'. Which of the following Snowflake SQL queries is the MOST efficient and statistically sound way to calculate the Pearson correlation coefficient between these two columns, handling potential NULL values appropriately?

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

Question 4

You've built a regression model in Snowflake to predict customer churn. You've calculated the R-squared score on your test data and found it to be 0.65. However, after deploying the model to production and monitoring its performance over several weeks, you notice the model's predictive accuracy has significantly decreased. Which of the following factors could contribute to this performance degradation?
Select all that apply.

A. Data drift: The distribution of the input features in the production data has changed significantly compared to the training data.
B. Feature engineering inconsistencies: The feature engineering steps applied to the production data are different from those applied during training.
C. Overfitting: The model learned the training data too well, capturing noise and specific patterns that do not generalize to new data.
D. Bias Variance trade off : Model is having high bias.
E. Increased data volume: The production data volume has increased significantly, causing resource contention and impacting model performance in Snowflake.

Question 5

You've built a model in Snowflake to predict the likelihood of a customer clicking on an advertisement. The model outputs a probability score between 0 and 1. You want to determine the optimal threshold to use for converting these probabilities into binary predictions (click/no-click). Your business stakeholders have provided the following information: Cost of showing an ad: $0.10; Revenue generated from a click: $1.00; You have access to a table 'AD_PREDICTIONS' with columns 'CUSTOMER_ID', 'PREDICTED_PROBABILITY' , and 'ACTUAL CLICK' (1 for click, 0 for no click). Which of the following approaches would be the MOST appropriate for selecting the optimal probability threshold to maximize profit, and why?

A. Select a very high probability threshold (e.g., 0.9) to ensure that only the most likely clicks are targeted, minimizing wasted ad spend.
B. Calculate the point on the ROC curve closest to the top-left corner (perfect classification) and use the corresponding threshold. This optimizes for both sensitivity and specificity.
C. Use the precision-recall curve to find the threshold that maximizes the F1 -score, balancing precision and recall.
D. Select a threshold of 0.5, as this is a common default threshold for binary classification problems.
E. Iterate through a range of probability thresholds (e.g., 0.01 to 0.99), and for each threshold, calculate the profit using SQL in Snowflake: 'SELECT SUM(CASE WHEN PREDICTED PROBABILITY threshold THEN CASE WHEN ACTUAL CLICK = 1 THEN 0.9 ELSE -0.1 END ELSE O END) AS Profit FROM AD_PREDICTIONS;' Choose the threshold that maximizes the profit.

Solutions: