AI Associate Data for AI Practice Test 5 — 15 Questions

Question 1

A Salesforce admin is preparing a dataset in Data Cloud to train an Einstein Discovery model that predicts customer churn. The dataset contains a field 'CustomerSince' with values in the format 'YYYY-MM-DD'. The admin wants to use this field as a feature but notices the model treats it as a string. What should the admin do to ensure the model can use the temporal information effectively?

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

Convert the 'CustomerSince' field to a numeric value representing the number of days since the customer joined.. For a churn prediction model, the tenure of a customer is a critical feature. Converting the join date to a numeric value like days since joining creates a continuous variable that the model can directly use to learn patterns. This transformation preserves the temporal information in a format that algorithms can process, unlike raw dates or partial extractions that lose granularity.

Answer

Leave the field as a string and rely on Einstein Discovery's automatic feature engineering to handle date types.

Answer

Change the field type to 'Date' in Data Cloud and use it directly as a feature.

Answer

Create a new field that extracts the year from 'CustomerSince' and use that as a feature.

Question 2

A Salesforce administrator is preparing the Data Cloud data model that will feed an Einstein Prediction Builder model for customer churn. The churn label is stored in a custom object field named Churn_Risk__c, which is a picklist with values High, Medium, and Low. The team wants the model to output a numeric probability for each customer. Which step should the administrator take before creating the model?

Accepted Answer

Create a Data Cloud stream and data model object (DMO) that maps the churn label and predictor fields, then configure the model in Einstein Prediction Builder using that DMO.. Einstein Prediction Builder consumes data from a Data Cloud data model object, so the churn label and predictors must be ingested via a stream and mapped to a DMO before the model can be created. Direct CSV uploads and CRM Analytics recipes do not satisfy that requirement, and collapsing the multi-class label into a binary flag would lose information the business needs.

Answer

Export Churn_Risk__c to a CSV file and upload it directly to Einstein Prediction Builder as a training dataset.

Answer

Enable Salesforce CRM Analytics and build a recipe that writes the churn label back to the Account object before training.

Answer

Convert Churn_Risk__c into a numeric binary field using a formula that maps High to 1 and all other values to 0.

Question 3

A Salesforce administrator is preparing a dataset for an Einstein Discovery model that predicts customer churn. The dataset includes a field called 'LastActivityDate' that is populated for only 40% of customer records. The administrator wants to handle this missing data to improve model accuracy. Which approach is most appropriate according to Salesforce best practices?

Accepted Answer

Create a binary indicator variable to flag missing 'LastActivityDate' and let Einstein Discovery handle the missing values automatically.. Einstein Discovery is designed to handle missing values automatically, but creating an explicit missing indicator can improve model performance when missingness is informative. Preserving all records avoids bias and data loss, while the indicator lets the model distinguish between truly absent activity and unknown values. This aligns with Salesforce's guidance to leverage built-in capabilities and engineer features that capture data quality nuances.

Answer

Remove all records with missing 'LastActivityDate' values from the dataset.

Answer

Replace missing 'LastActivityDate' values with a fixed date, such as 01/01/1900.

Answer

Impute missing values using the mean or median 'LastActivityDate' from the non-missing records.

Question 4

A data scientist is preparing a dataset for an Einstein Discovery model to predict sales opportunity win rates. The dataset contains several fields, including 'OpportunityAmount', 'CloseDate', 'StageName', and 'Probability'. The scientist wants to ensure the model is trained on high-quality data. Which two data preparation steps are most critical to improve model accuracy? (Choose two.)

Accepted Answer

Verify that 'Probability' values are within the valid range of 0 to 100.. Ensuring consistent date formats and valid probability ranges are critical data quality checks. Inconsistent dates can lead to misinterpretation of temporal patterns, while invalid probabilities introduce noise and incorrect relationships. These steps directly improve the reliability of the features used by the model, leading to better predictions.

Answer

Handle missing values in 'OpportunityAmount' by imputing them with the median amount.

Answer

Convert 'StageName' into a numeric value using label encoding.

Answer

Remove outliers in 'OpportunityAmount' that are beyond three standard deviations from the mean.

Question 5

A financial services company uses Salesforce Data Cloud to build a predictive model for loan default risk. The data includes a field 'Credit_Score' with values ranging from 300 to 850. The data engineer notices that the distribution is heavily skewed, with most scores above 700. The model's performance is poor, and the engineer suspects that the skewed distribution is causing issues. Which technique should the engineer apply to address this?

Accepted Answer

Use a Box-Cox transformation to stabilize variance and make the data more normal.. Skewed numerical features can hinder model performance because many algorithms assume normal distribution. The Box-Cox transformation is a powerful method to reduce skewness and stabilize variance, making the data more suitable for modeling. It is particularly effective for positive, continuous data like credit scores. Applying Box-Cox can lead to better model convergence and accuracy.

Answer

Apply a logarithmic transformation to the credit scores.

Answer

Discretize the credit scores into bins and treat them as categorical.

Answer

Normalize the credit scores to a 0-1 range using min-max scaling.

Question 6

A Salesforce admin is preparing data for an Einstein Discovery model to predict customer lifetime value. The dataset contains a 'CustomerSince' date field. Which data quality issue is most critical to address before training the model?

Accepted Answer

The 'CustomerSince' field has inconsistent date formats.. Inconsistent date formats are a critical data quality issue because they can prevent the field from being parsed correctly, leading to errors or loss of information. Standardizing formats ensures the field can be used as a reliable predictor. Other issues like missing values or future dates are important but secondary to format consistency.

Answer

The 'CustomerSince' field has missing values for some customers.

Answer

The 'CustomerSince' field is stored as a text string instead of a date type.

Answer

The 'CustomerSince' field contains some future dates.

Question 7

A Salesforce admin is preparing data for an Einstein Discovery model to predict opportunity win rates. The dataset includes fields such as 'Amount', 'CloseDate', 'StageName', and 'LeadSource'. Which two data quality dimensions are most critical to validate before training the model? (Choose two.)

Accepted Answer

Completeness. Completeness and consistency are critical for preparing data for an Einstein Discovery model. Missing values can bias predictions, and inconsistent categorical values can confuse the model. While other dimensions like timeliness, uniqueness, and validity are important, they are not as directly impactful for this scenario.

Answer

Uniqueness

Answer

Timeliness

Answer

Validity

Question 8

A Salesforce admin is preparing a dataset in Data Cloud to train an Einstein Discovery model that predicts customer churn. The dataset contains a field 'CustomerSince' with values in the format 'YYYY-MM-DD'. The admin wants to use this field as a feature but notices the model treats it as a string. What should the admin do to ensure the model can use the temporal information effectively?

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

Apply one-hot encoding to create binary columns for each lead source.. One-hot encoding is the correct technique for nominal categorical variables like LeadSource. It creates binary indicators that allow the model to treat each category separately without implying order. This improves the model's ability to detect patterns associated with each lead source, leading to better predictions.

Answer

Remove the 'LeadSource' field because it is categorical and not useful for prediction.

Answer

Group all lead sources except 'Web' into an 'Other' category to reduce cardinality.

Answer

Convert 'LeadSource' into a numeric field by assigning each source a unique integer.

Question 9

A Salesforce admin is integrating external data from an ERP system into Data Cloud to train an AI model. The external data contains customer order history with fields like OrderDate, ProductID, and Quantity. The admin needs to ensure the data is properly mapped and unified. Which Data Cloud feature should they use to define the relationship between the ERP data and existing Customer 360 data?

Accepted Answer

Identity Resolution. Identity Resolution in Data Cloud is specifically designed to unify records from multiple sources by defining match rules. It creates a unified individual profile that consolidates data from ERP and other systems. This unified profile is then used by AI models to access a complete view of the customer, which is essential for accurate predictions.

Answer

Data Streams

Answer

Data Mapping

Answer

Data Model Objects

Question 10

A Salesforce admin is preparing a dataset in Data Cloud to train an Einstein Discovery model that predicts customer churn. The dataset contains a field 'CustomerSince' with values in the format 'YYYY-MM-DD'. The admin wants to use this field as a feature but notices the model treats it as a string. What should the admin do to ensure the model can use the temporal information effectively?

Choose an answer to begin — your selection is scored in the full session.

Accepted Answer

The model is biased because 'TotalSpend' has a larger scale; feature scaling should be applied to bring all features to a similar range.. Feature scaling is a common preprocessing step that ensures all features contribute equally to the model. Without scaling, features with larger magnitudes, like 'TotalSpend', can dominate the learning process, leading to biased feature importance. Applying techniques like min-max normalization or z-score standardization brings all features to a comparable scale, allowing the model to accurately assess each feature's predictive value.

Answer

The model is overfitting because it relies too much on one feature; remove 'TotalSpend' from the dataset.

Answer

The model is correctly identifying that 'TotalSpend' is more predictive; no action is needed.

Answer

The model is underfitting because it cannot capture the relationship; add more features to the dataset.

Question 11

A Salesforce administrator is preparing a dataset for an AI model that predicts customer churn. The dataset contains missing values in several fields, including 'Last_Purchase_Date' and 'Customer_Support_Calls'. The administrator needs to handle these missing values appropriately to avoid degrading model performance. Which two actions should the administrator take? (Choose two.)

Accepted Answer

Impute missing values in 'Customer_Support_Calls' with the mean of the column.. Handling missing data is critical for model accuracy. Simple mean imputation for numerical fields like support calls is effective when data is missing at random. More sophisticated methods like KNN imputation can capture complex patterns and are suitable for both numerical and date fields, though they require careful implementation. Removing rows or using single-value imputation for dates can introduce bias or lose information, so these should be avoided when better options exist.

Answer

Impute missing 'Last_Purchase_Date' values with the most recent purchase date in the dataset.

Answer

For 'Last_Purchase_Date', create a new binary feature indicating whether the date is missing.

Answer

Remove all rows with any missing values to ensure data completeness.

Question 12

A data engineer is preparing a dataset for an Einstein Discovery model to predict customer churn. The dataset includes fields such as 'CustomerID', 'LastPurchaseDate', 'TotalSpend', and 'SupportTickets'. The engineer needs to ensure the data is properly formatted and cleaned. Which two actions are essential for preparing the data for the model? (Choose two.)

Accepted Answer

Check for and handle missing values in 'TotalSpend' and 'SupportTickets'.. Ensuring a unique and complete 'CustomerID' is essential for correctly linking records to customers. Handling missing values in key numerical fields like 'TotalSpend' and 'SupportTickets' prevents errors and bias. These two actions are fundamental to preparing clean data for an Einstein Discovery churn model.

Answer

Remove all records where 'SupportTickets' is greater than zero.

Answer

Create a new field that concatenates 'CustomerID' and 'LastPurchaseDate'.

Answer

Convert 'LastPurchaseDate' to a numeric value representing days since last purchase.

Question 13

A company wants to use Einstein Prediction Builder to predict customer churn. Which data preparation step is essential before building the model?

Accepted Answer

Define the prediction objective and the target date field.. Einstein Prediction Builder requires you to define the prediction objective (e.g., 'Will this customer churn?') and specify the target date field that marks the event. This step is essential as it tells the model what to predict and over what time window, enabling the automated feature engineering and model training process.

Answer

Ensure the data is in a Salesforce connected data source like Data Cloud.

Answer

Create a formula field to calculate the churn probability.

Answer

Create a new custom object to store the prediction results.

Question 14

A data scientist needs to prepare data for Einstein Discovery. The dataset includes a field 'Customer_Status__c' with values 'Active', 'Inactive', and 'Churned'. How should this field be treated?

Accepted Answer

Keep as a text field and let Einstein Discovery handle it as a categorical predictor.. Einstein Discovery natively supports text fields as categorical predictors, automatically encoding them for model training. The platform handles string values like 'Active', 'Inactive', and 'Churned' without requiring manual transformation, preserving the semantic meaning and cardinality of the data.

Answer

Create separate boolean fields for each value to improve model accuracy.

Answer

Remove the field because text fields cannot be used in Einstein Discovery.

Answer

Convert to numeric values 1, 2, 3 to preserve order.

Question 15

A Salesforce admin is training an Einstein Bot to answer customer questions. Which data source should the bot use to provide accurate responses?

Accepted Answer

Knowledge articles with a published status.. Knowledge articles with a published status are the correct data source because they contain curated, approved, and structured information that Einstein Bot can reliably use to generate accurate responses. The bot leverages natural language processing to match customer questions against these articles, ensuring answers are based on verified content rather than unstructured or transient data.

Answer

Chatter posts from the product team.

Answer

Case records from the last 30 days.

Answer

Lead and contact reports.