Predictive Analytics Fundamentals — Quiz

Answer all 12 questions, then submit. You need 70% to pass. Log in to save progress.

Question 1
Predicting next quarter's revenue (a number) is a…
A classification task
B regression task
C clustering task
D scraping task
Question 2
Why split data into training and test sets?
A To save memory
B To judge the model on data it has not seen
C Because Pandas requires it
D To make training faster
Question 3
In linear regression, the slope (coefficient) tells you…
A the number of rows
B the effect of one unit of a feature on the target
C the R-squared
D the error
Question 4
An R-squared of 0.87 means the model…
A is 87% accurate at classification
B explains 87% of the variation in the target
C has 87% error
D used 87 features
Question 5
Which model gives the probability of churn for each customer?
A LinearRegression
B LogisticRegression with predict_proba
C describe()
D groupby
Question 6
Which model is most naturally explainable as yes/no rules?
A a neural network
B a decision tree
C linear regression
D k-means
Question 7
On a dataset where only 2% churn, accuracy is misleading because…
A it is too slow
B predicting 'no churn' for everyone scores ~98% yet is useless
C it needs more columns
D accuracy is not a metric
Question 8
Recall measures…
A how fast the model runs
B of all real positives, how many the model caught
C the number of features
D the R-squared
Question 9
Precision measures…
A of those predicted positive, how many were actually positive
B all the data
C the slope
D the training time
Question 10
Feature importance is valuable because it…
A makes the model bigger
B shows which factors drive the outcome so the business can act
C removes the need for data
D guarantees accuracy
Question 11
Which technique forecasts a time series with trend and seasonality?
A a bar chart
B exponential smoothing / Holt-Winters
C a pivot table
D a join
Question 12
What is the analyst's most important habit with any model?
A use the most complex model possible
B translate results into a clear, actionable recommendation
C hide the uncertainty
D never test it