,Exam Blueprint & Instructions
WGU D600 Task 2: Logistic Regression Analysis | Complete Report, Python Modeling, Confusion Matrix & Results
Part I · Exam Questions (1-200)
1. Logistic Regression Fundamentals, Odds & Logit Function (22 q) ....5
2. Data Preprocessing, Encoding & Scaling for Logistic Models (22 q) ....6
3. Python Implementation: LogisticRegression & Solver Selection (22 q) ....7
4. Confusion Matrix, Accuracy, Precision, Recall & F1-Score (23 q) ....8
5. ROC Curve, AUC, Threshold Tuning & Probability Calibration (22 q) ....9
6. Model Evaluation, Cross-Validation & Regularization (L1/L2) (22 q) . . . . 10
7. WGU D600 Task 2 Requirements, Report Structure & Documentation (22 q) . . . . 11
8. Handling Imbalanced Data, Class Weights, SMOTE & Sampling (22 q) . . . . 12
9. Business Implications, Limitations, Ethics & Recommendations (23 q) . . . . 13
Instructions:
• 200 original multiple-choice questions for WGU D600 Task 2 mastery.
• Covers logistic regression, confusion matrix, ROC/AUC, threshold tuning, imbalanced data handling.
• All code examples use sklearn, pandas, and WGU-compliant documentation standards.
• Correct answers and rationales follow each item immediately.
• Original content, not actual WGU assessment, for study only.
,SECTION ONE: QUESTIONS 1-200
Choose the single best answer. Correct answers and rationales follow each item.
Logistic Regression Fundamentals, Odds & Logit Function
Question 1. In WGU D600 Task 2, what is the primary purpose of plotting ROC curve and
calculating AUC to assess discriminative ability of logistic model?
A. Omit WGU D600 documentation of code, assumptions, visualizations, and limitations from
report
B. Use accuracy alone on imbalanced data without confusion matrix or recall analysis
C. Plot ROC curve (FPR vs TPR) and compute AUC; tune threshold using precision_recall_curve per
business cost
D. Ignore threshold tuning and report only default 0.5 probability predictions without
calibration
CORRECT ANSWER: C. Plot ROC curve (FPR vs TPR) and compute AUC; tune threshold using
precision_recall_curve per business cost
RATIONALE: For WGU D600 Task 2 Logistic Regression, plotting ROC curve and calculating AUC to assess
discriminative ability of logistic model must be handled with correct methodology. Correct option
(C) ensures: Plot ROC curve (FPR vs TPR) and compute AUC; tune threshold using
precision_recall_curve per business cost. This aligns with logistic regression theory (sigmoid,
maximum likelihood), WGU rubric (Python modeling, confusion matrix, results, documentation), and
ethical business reporting. Incorrect options cause data leakage, misleading metrics, or rubric
failure.
Question 2. When implementing interpreting odds ratio and logit function in logistic
regression for binary classification in Python, which approach avoids common errors?
A. Ignore threshold tuning and report only default 0.5 probability predictions without
calibration
B. Omit WGU D600 documentation of code, assumptions, visualizations, and limitations from
report
C. Correct WGU Task 2 practice: interpreting odds ratio and logit function in logistic
regression for binary cla with proper Python modeling, rubric-compliant documentation, and
business translation
D. Fit preprocessing including SMOTE and scaling on full dataset before train-test split
causing leakage
CORRECT ANSWER: C. Correct WGU Task 2 practice: interpreting odds ratio and logit function
in logistic regression for binary cla with proper Python modeling, rubric-compliant
documentation, and business translation
RATIONALE: For WGU D600 Task 2 Logistic Regression, interpreting odds ratio and logit function in
logistic regression for binary classification must be handled with correct methodology. Correct
option (C) ensures: Correct WGU Task 2 practice: interpreting odds ratio and logit function in
logistic regression for binary cla with proper Python modeling, rubric-compliant documentation, and
business translation. This aligns with logistic regression theory (sigmoid, maximum likelihood), WGU
rubric (Python modeling, confusion matrix, results, documentation), and ethical business reporting.
Incorrect options cause data leakage, misleading metrics, or rubric failure.
, WGU D600 Task 2 - Logistic Regression Fundamentals, Odds & Logit Function
Question 3. A student observes applying class_weight='balanced' or SMOTE to address class
imbalance in training data with low recall. What is the best next step?
A. Omit WGU D600 documentation of code, assumptions, visualizations, and limitations from
report
B. Ignore threshold tuning and report only default 0.5 probability predictions without
calibration
C. Fit preprocessing including SMOTE and scaling on full dataset before train-test split
causing leakage
D. Use class_weight='balanced' or SMOTE on training set only, evaluate with stratified CV and
PR-AUC, not just accuracy
CORRECT ANSWER: D. Use class_weight='balanced' or SMOTE on training set only, evaluate with
stratified CV and PR-AUC, not just accuracy
RATIONALE: For WGU D600 Task 2 Logistic Regression, applying class_weight='balanced' or SMOTE to
address class imbalance in training data must be handled with correct methodology. Correct option
(D) ensures: Use class_weight='balanced' or SMOTE on training set only, evaluate with stratified CV
and PR-AUC, not just accuracy. This aligns with logistic regression theory (sigmoid, maximum
likelihood), WGU rubric (Python modeling, confusion matrix, results, documentation), and ethical
business reporting. Incorrect options cause data leakage, misleading metrics, or rubric failure.
Question 4. Which Python code correctly implements tuning classification threshold from
0.5 to improve recall for imbalanced business problem?
A. Use class_weight='balanced' or SMOTE on training set only, evaluate with stratified CV and
PR-AUC, not just accuracy
B. Fit preprocessing including SMOTE and scaling on full dataset before train-test split
causing leakage
C. Ignore threshold tuning and report only default 0.5 probability predictions without
calibration
D. Omit WGU D600 documentation of code, assumptions, visualizations, and limitations from
report
CORRECT ANSWER: A. Use class_weight='balanced' or SMOTE on training set only, evaluate with
stratified CV and PR-AUC, not just accuracy
RATIONALE: For WGU D600 Task 2 Logistic Regression, tuning classification threshold from 0.5 to
improve recall for imbalanced business problem must be handled with correct methodology. Correct
option (A) ensures: Use class_weight='balanced' or SMOTE on training set only, evaluate with
stratified CV and PR-AUC, not just accuracy. This aligns with logistic regression theory (sigmoid,
maximum likelihood), WGU rubric (Python modeling, confusion matrix, results, documentation), and
ethical business reporting. Incorrect options cause data leakage, misleading metrics, or rubric
failure.
Question 5. What metric is most informative when evaluating comparing logistic regression
to decision trees for interpretability in D600 business context?
A. Use accuracy alone on imbalanced data without confusion matrix or recall analysis
B. Ignore threshold tuning and report only default 0.5 probability predictions without
calibration
C. Omit WGU D600 documentation of code, assumptions, visualizations, and limitations from
report
WGU D600 Task 2: Logistic Regression Analysis | Complete Report, Python Modeling, Confusion Matrix & Results
Part I · Exam Questions (1-200)
1. Logistic Regression Fundamentals, Odds & Logit Function (22 q) ....5
2. Data Preprocessing, Encoding & Scaling for Logistic Models (22 q) ....6
3. Python Implementation: LogisticRegression & Solver Selection (22 q) ....7
4. Confusion Matrix, Accuracy, Precision, Recall & F1-Score (23 q) ....8
5. ROC Curve, AUC, Threshold Tuning & Probability Calibration (22 q) ....9
6. Model Evaluation, Cross-Validation & Regularization (L1/L2) (22 q) . . . . 10
7. WGU D600 Task 2 Requirements, Report Structure & Documentation (22 q) . . . . 11
8. Handling Imbalanced Data, Class Weights, SMOTE & Sampling (22 q) . . . . 12
9. Business Implications, Limitations, Ethics & Recommendations (23 q) . . . . 13
Instructions:
• 200 original multiple-choice questions for WGU D600 Task 2 mastery.
• Covers logistic regression, confusion matrix, ROC/AUC, threshold tuning, imbalanced data handling.
• All code examples use sklearn, pandas, and WGU-compliant documentation standards.
• Correct answers and rationales follow each item immediately.
• Original content, not actual WGU assessment, for study only.
,SECTION ONE: QUESTIONS 1-200
Choose the single best answer. Correct answers and rationales follow each item.
Logistic Regression Fundamentals, Odds & Logit Function
Question 1. In WGU D600 Task 2, what is the primary purpose of plotting ROC curve and
calculating AUC to assess discriminative ability of logistic model?
A. Omit WGU D600 documentation of code, assumptions, visualizations, and limitations from
report
B. Use accuracy alone on imbalanced data without confusion matrix or recall analysis
C. Plot ROC curve (FPR vs TPR) and compute AUC; tune threshold using precision_recall_curve per
business cost
D. Ignore threshold tuning and report only default 0.5 probability predictions without
calibration
CORRECT ANSWER: C. Plot ROC curve (FPR vs TPR) and compute AUC; tune threshold using
precision_recall_curve per business cost
RATIONALE: For WGU D600 Task 2 Logistic Regression, plotting ROC curve and calculating AUC to assess
discriminative ability of logistic model must be handled with correct methodology. Correct option
(C) ensures: Plot ROC curve (FPR vs TPR) and compute AUC; tune threshold using
precision_recall_curve per business cost. This aligns with logistic regression theory (sigmoid,
maximum likelihood), WGU rubric (Python modeling, confusion matrix, results, documentation), and
ethical business reporting. Incorrect options cause data leakage, misleading metrics, or rubric
failure.
Question 2. When implementing interpreting odds ratio and logit function in logistic
regression for binary classification in Python, which approach avoids common errors?
A. Ignore threshold tuning and report only default 0.5 probability predictions without
calibration
B. Omit WGU D600 documentation of code, assumptions, visualizations, and limitations from
report
C. Correct WGU Task 2 practice: interpreting odds ratio and logit function in logistic
regression for binary cla with proper Python modeling, rubric-compliant documentation, and
business translation
D. Fit preprocessing including SMOTE and scaling on full dataset before train-test split
causing leakage
CORRECT ANSWER: C. Correct WGU Task 2 practice: interpreting odds ratio and logit function
in logistic regression for binary cla with proper Python modeling, rubric-compliant
documentation, and business translation
RATIONALE: For WGU D600 Task 2 Logistic Regression, interpreting odds ratio and logit function in
logistic regression for binary classification must be handled with correct methodology. Correct
option (C) ensures: Correct WGU Task 2 practice: interpreting odds ratio and logit function in
logistic regression for binary cla with proper Python modeling, rubric-compliant documentation, and
business translation. This aligns with logistic regression theory (sigmoid, maximum likelihood), WGU
rubric (Python modeling, confusion matrix, results, documentation), and ethical business reporting.
Incorrect options cause data leakage, misleading metrics, or rubric failure.
, WGU D600 Task 2 - Logistic Regression Fundamentals, Odds & Logit Function
Question 3. A student observes applying class_weight='balanced' or SMOTE to address class
imbalance in training data with low recall. What is the best next step?
A. Omit WGU D600 documentation of code, assumptions, visualizations, and limitations from
report
B. Ignore threshold tuning and report only default 0.5 probability predictions without
calibration
C. Fit preprocessing including SMOTE and scaling on full dataset before train-test split
causing leakage
D. Use class_weight='balanced' or SMOTE on training set only, evaluate with stratified CV and
PR-AUC, not just accuracy
CORRECT ANSWER: D. Use class_weight='balanced' or SMOTE on training set only, evaluate with
stratified CV and PR-AUC, not just accuracy
RATIONALE: For WGU D600 Task 2 Logistic Regression, applying class_weight='balanced' or SMOTE to
address class imbalance in training data must be handled with correct methodology. Correct option
(D) ensures: Use class_weight='balanced' or SMOTE on training set only, evaluate with stratified CV
and PR-AUC, not just accuracy. This aligns with logistic regression theory (sigmoid, maximum
likelihood), WGU rubric (Python modeling, confusion matrix, results, documentation), and ethical
business reporting. Incorrect options cause data leakage, misleading metrics, or rubric failure.
Question 4. Which Python code correctly implements tuning classification threshold from
0.5 to improve recall for imbalanced business problem?
A. Use class_weight='balanced' or SMOTE on training set only, evaluate with stratified CV and
PR-AUC, not just accuracy
B. Fit preprocessing including SMOTE and scaling on full dataset before train-test split
causing leakage
C. Ignore threshold tuning and report only default 0.5 probability predictions without
calibration
D. Omit WGU D600 documentation of code, assumptions, visualizations, and limitations from
report
CORRECT ANSWER: A. Use class_weight='balanced' or SMOTE on training set only, evaluate with
stratified CV and PR-AUC, not just accuracy
RATIONALE: For WGU D600 Task 2 Logistic Regression, tuning classification threshold from 0.5 to
improve recall for imbalanced business problem must be handled with correct methodology. Correct
option (A) ensures: Use class_weight='balanced' or SMOTE on training set only, evaluate with
stratified CV and PR-AUC, not just accuracy. This aligns with logistic regression theory (sigmoid,
maximum likelihood), WGU rubric (Python modeling, confusion matrix, results, documentation), and
ethical business reporting. Incorrect options cause data leakage, misleading metrics, or rubric
failure.
Question 5. What metric is most informative when evaluating comparing logistic regression
to decision trees for interpretability in D600 business context?
A. Use accuracy alone on imbalanced data without confusion matrix or recall analysis
B. Ignore threshold tuning and report only default 0.5 probability predictions without
calibration
C. Omit WGU D600 documentation of code, assumptions, visualizations, and limitations from
report
