Free SOA Exam PA (Predictive Analytics) Tree-Based Models Practice Questions
Tree-based models on SOA Exam PA cover decision trees, random forests, and gradient boosting machines: reading fitted trees and variable importance plots, tuning parameters, and weighing interpretability against predictive performance.
75 Questions
16 Easy
52 Medium
7 Hard
2026 Syllabus
Sample Questions
Question 1
Easy
Which statement best describes the fundamental difference between boosting and bagging for tree ensembles?
🎉
Correct Answer: B
Solution
B is correct. Boosting is a sequential procedure: each new (typically shallow) tree is fit to correct the mistakes of the ensemble built so far, chiefly reducing bias. Bagging fits many independent trees on bootstrap resamples and averages them, chiefly reducing variance.
Question 2
Medium
In rpart.control, which statement correctly matches a control parameter to its effect on tree growth?
🎉
Correct Answer: C
Solution
C is correct. minbucket sets the minimum number of observations permitted in a leaf. cp is the minimum improvement in the (scaled) fit required to make a split, minsplit is the minimum node size before a split is attempted, and maxdepth caps tree depth. Lowering cp permits more splits, producing a larger, not smaller, tree.
Question 3
Hard
A GBM and a random forest are both fit to the same claims dataset. Which statement about their bias-variance behavior is most accurate?
🎉
Correct Answer: C
Solution
C is correct. Random forests average many decorrelated high-variance trees, so their gain is variance reduction. Boosting sequentially reduces bias by targeting residuals, and because it keeps adding corrective trees it can overfit when the learning rate, depth, or number of rounds are not tuned.
FreeFellow is a free exam prep library for actuarial (SOA & CAS), CFA, CFP, CPA, CAIA, GARP FRM, IRS Enrolled Agent, IMA CMA, and FINRA / NASAA securities licensing candidates. The entire question bank, written solutions, and lessons are free for every candidate, with no trial period and no credit card. Lessons include narrated audio, and every constructed-response item has a copy-to-AI prompt builder so candidates can paste their answer into their own ChatGPT or Claude for self-graded feedback; Fellow members get instant AI grading on essays against the official rubric (currently CFA Level III, expanding to other essay-bearing sections).
The 70% you need to pass (question bank, written solutions, lessons, formula sheet, mixed practice, readiness tracking) is free forever, with no trial period and no credit card. Become a Fellow ($59/quarter or $149/year per track) to unlock mock exams, flashcards with spaced repetition, performance analytics, AI essay grading, and a personalized study plan.