Question 1
A lender wants a fraud model, but leaders cannot state what decision the output will support. What should happen first?
Show answer & explanation
Correct answer: D - Define the decision and success criteria
10 exam-style questions with answers and explanations, straight from our 1,030-question bank. Tap an answer to check yourself. When you're ready, take the scored version in the free practice test.
These 10 free CDS questions are organized by exam domain, so you can see how each part of the Certified Data Scientist - Associate/Practitioner blueprint is tested. Reveal the answer and explanation under each question.
A lender wants a fraud model, but leaders cannot state what decision the output will support. What should happen first?
Correct answer: D - Define the decision and success criteria
A system receives millions of events every second. Which Big Data characteristic describes the arrival speed?
Correct answer: C - Velocity
A salary dataset contains a few extremely high executive salaries. Which measure of location is most resistant to those extreme values?
Correct answer: A - Median
An analyst inherits a pipeline with undocumented transformations. Which improvement best supports repeatable analysis?
Correct answer: D - Version the transformation code and settings
A manager asks which customers will cancel next month. What type of analytics is requested?
Correct answer: A - Predictive analytics
A customer appears twice because systems use different identifiers. Which data quality issue is present?
Correct answer: B - Uniqueness
Exploration identifies unusually large sales contracts. Source checks confirm they are valid transactions in the population being analyzed. What is the best next action?
Correct answer: A - Retain them and assess their influence on results
A model intended to support a loan decision uses a feature first recorded after that decision. Why might its retrospective test score overstate deployable performance?
Correct answer: D - Information unavailable at prediction time leaks into the model
A company wants a supervised model of whether a subscriber will cancel, using a binary yes-or-no outcome. Which listed method directly models a binary outcome probability?
Correct answer: D - Logistic regression
In an observational study, wellness-app users have better health outcomes than nonusers. Users also tend to exercise more before installing the app. What most directly limits attributing the difference to the app?
Correct answer: C - Pre-existing exercise habits may confound the association
The CDS exam also covers these domains. Drill them in the full free practice test:
The full bank has 1,020 more CDS questions with explanations.
Continue in the free practice test →
View plans