PEGACPDS26V1 Certified Pega Data Scientist (PEGACPDS26V1) Practice Questions
Prepare for PEGACPDS26V1 with more than an answer.
- Exam fee
- $175 USD
- Time limit
- 90 minutes
- Questions on the exam
- 50
- Passing score
- 70%
- Level
- Beginner
- Valid for
- Tied to the product version; Pega recommends recertifying/upskilling with each new Infinity release
Domains covered on the exam 7
- AI for Customer Decision Hub6%
- Adaptive Analytics28%
- Predictive Analytics20%
- Prediction Patterns18%
- Governance4%
- Pega Process AI10%
- Pega NLP14%
- 1
Which underlying machine learning algorithm powers the Adaptive Decision Manager (ADM) in Pega Infinity '26 to enable real-time, self-learning predictions?
Show answer details
Correct answer: C
Pega's Adaptive Decision Manager (ADM) primarily uses a proprietary implementation based on the Naive Bayes algorithm. This algorithm is highly efficient for real-time, incremental learning because it calculates probabilities based on predictor bins and updates these probabilities instantly as new responses are recorded.
- 2
A data scientist is analyzing the performance of a specific adaptive model using the monitor tab in Prediction Studio. They notice that a predictor named 'Customer_Age' has a calculated AUC of 49%. What does this AUC value indicate about the predictor's ability to differentiate between positive and negative outcomes?
Show answer details
Correct answer: C
In ROC analysis, an Area Under the Curve (AUC) of 50% represents a random guess. A value of 49% indicates that the predictor has virtually no predictive power. In Pega ADM, predictors with an AUC below the activation threshold (typically 52%) remain inactive and do not influence the model's final propensity score.
- 3
In Pega's Adaptive Decision Manager, when does a candidate predictor transition from 'Inactive' to 'Active' status and begin contributing to the final propensity score?
Show answer details
Correct answer: A
ADM constantly evaluates all configured predictors. A predictor remains 'Inactive' until its predictive performance (measured by AUC) exceeds a specific activation threshold, which defaults to 52%. Once it crosses this threshold, it becomes 'Active' and its bins are used in calculating the final propensity.
- 4
When creating a new adaptive model from scratch in Prediction Studio, which TWO configuration elements are strictly required to define the boundaries of what the model will learn? (Select TWO)
Show answer details
Correct answer: A, C
To configure an adaptive model, you must supply it with Predictors (the input variables like demographics or behavior that it will use to find patterns) and Outcomes (the definition of positive and negative behaviors it is trying to predict). Without these two elements, the model has no data to learn from and no goal to achieve.
To configure an adaptive model, you must supply it with Predictors (the input variables like demographics or behavior that it will use to find patterns) and Outcomes (the definition of positive and negative behaviors it is trying to predict). Without these two elements, the model has no data to learn from and no goal to achieve.
- 5
A data scientist is reviewing the Next-Best-Action Designer arbitration configuration for a telecommunications company using Pega Customer Decision Hub '26. The company wants to ensure that a retention offer is prioritized over a cross-sell offer when a customer is highly likely to churn, even if the cross-sell offer has a higher historical conversion rate. Which component of the arbitration formula must the data scientist adjust to achieve this specific outcome?
Show answer details
Correct answer: B
Business levers allow you to apply business logic to prioritize certain actions over others under specific conditions. By applying a business lever (such as a weighting factor when a high churn propensity is detected), the arbitration formula will artificially boost the priority of the retention offer over the cross-sell offer, satisfying the business requirement.
- 6
An organization is deploying the out-of-the-box Predict Web Propensity model in Pega Customer Decision Hub. The marketing team notices that a small percentage of web visitors are not receiving the AI-driven next best actions, but instead receive a random selection of actions. What is the primary purpose of this behavior?
Show answer details
Correct answer: B
In Pega CDH, a control group is a small, randomly selected audience segment that receives random actions rather than AI-optimized next best actions. This allows the organization to measure the effectiveness (lift) of the AI model by comparing the response rates of the target group against the control group.
