DP-100 · Question #4
DP-100 Question #4: Real Exam Question with Answer & Explanation
The correct answer is A: Set the threshold to 0.5 and retrain if weighted Kappa deviates +/- 5% from 0.45.. {"question_number": 3, "question": "You need to implement a new cost factor scenario for the ad response models as illustrated in the performance curve exhibit. Which technique should you use?", "correct_answer": "A. Set the threshold to 0.5 and retrain if weighted Kappa deviates
Question
You need to implement a new cost factor scenario for the ad response models as illustrated in the performance curve exhibit. Which technique should you use?
Options
- ASet the threshold to 0.5 and retrain if weighted Kappa deviates +/- 5% from 0.45.
- BSet the threshold to 0.05 and retrain if weighted Kappa deviates +/- 5% from 0.5.
- CSet the threshold to 0.2 and retrain if weighted Kappa deviates +/- 5% from 0.6.
- DSet the threshold to 0.75 and retrain if weighted Kappa deviates +/- 5% from 0.15.
Explanation
{"question_number": 3, "question": "You need to implement a new cost factor scenario for the ad response models as illustrated in the performance curve exhibit. Which technique should you use?", "correct_answer": "A. Set the threshold to 0.5 and retrain if weighted Kappa deviates +/- 5% from 0.45.", "explanation": "A classification threshold of 0.5 is the standard balanced default that minimizes misclassification without skewing toward false positives or false negatives. Weighted Kappa measures agreement between predicted and actual categories while accounting for ordinal ordering - a baseline of 0.45 indicates moderate agreement, which is realistic for ad response models. Monitoring with a +/- 5% tolerance window triggers retraining only when model performance meaningfully degrades, avoiding unnecessary retraining cycles. The other options use thresholds (0.05, 0.2, 0.75) or Kappa baselines that do not align with a balanced cost scenario.", "generated_by": "claude-sonnet", "llm_judge_score": 3}
Topics
Community Discussion
No community discussion yet for this question.