WGU D499 MACHINE LEARNING PROJECT
1: FINDING DONORS FOR CHARITYML
[QUESTIONS 1-100] AND ANSWERS
UPDATED 2026/2027 | DETAILED
RATIONALES
Study-use notice: This is an original practice question bank based on the concepts
and skills associated with the CharityML donor-classification project. It is not an
official WGU exam, assessment, answer key, or confidential course material, and
no score or pass outcome is guaranteed.
INTRODUCTION
This advanced practice bank focuses on the technical reasoning required for a
supervised machine-learning project involving donor classification, preprocessing,
model selection, hyperparameter tuning, evaluation, and interpretation. The scenarios
emphasize the kinds of decisions a learner must make when working with
demographic and socioeconomic data to predict whether an individual is likely to
exceed a specified income threshold. Questions cover exploratory data analysis,
categorical encoding, feature scaling, train/test splitting, baseline models, decision
trees, ensemble methods, support vector machines, logistic regression, GridSearchCV,
cross-validation, precision, recall, F0.5, confusion matrices, class imbalance, feature
importance, and model selection. Several questions require distinguishing between a
model that achieves strong overall accuracy and one that better satisfies the project's
business objective. The bank also emphasizes avoiding data leakage, interpreting
preprocessing correctly, explaining model behavior, and selecting an appropriate final
estimator. The goal is to develop practical reasoning rather than memorization.
CORE DOMAINS TESTED
1.
CharityML donor-classification problem formulation
2.
3.
Exploratory data analysis and preprocessing
4.
5.
1
,Categorical feature encoding
6.
7.
Numerical feature handling and scaling
8.
9.
Train/test splitting and reproducibility
10.
11.
Baseline and supervised learning models
12.
13.
Decision trees and ensemble methods
14.
15.
Support Vector Machines and Logistic Regression
16.
17.
GridSearchCV and hyperparameter optimization
18.
19.
Cross-validation and model selection
20.
21.
Accuracy, precision, recall, F-score, and F0.5
22.
23.
Confusion-matrix interpretation
24.
25.
2
, Class imbalance
26.
27.
Feature importance and model interpretation
28.
29.
Data leakage and pipeline design
30.
31.
Business-oriented model evaluation
32.
33.
Model efficiency and computational tradeoffs
34.
35.
Final-model justification
36.
QUESTIONS 1-100
Q1: What is the primary supervised-learning objective in the
CharityML donor-classification project?
A) Predict the exact annual income of every individual
B) Predict whether an individual belongs to the higher-income class based on
available features
C) Cluster individuals into demographic groups without labels
D) Reduce the dimensionality of the census dataset
Rationale: The correct answer is B because the task is a binary classification problem
in which the model predicts membership in the target income category. A is incorrect
because exact income prediction would be regression. C describes unsupervised
learning rather than supervised classification. D describes dimensionality reduction,
which may be useful during preprocessing but is not the project's primary objective.
3
1: FINDING DONORS FOR CHARITYML
[QUESTIONS 1-100] AND ANSWERS
UPDATED 2026/2027 | DETAILED
RATIONALES
Study-use notice: This is an original practice question bank based on the concepts
and skills associated with the CharityML donor-classification project. It is not an
official WGU exam, assessment, answer key, or confidential course material, and
no score or pass outcome is guaranteed.
INTRODUCTION
This advanced practice bank focuses on the technical reasoning required for a
supervised machine-learning project involving donor classification, preprocessing,
model selection, hyperparameter tuning, evaluation, and interpretation. The scenarios
emphasize the kinds of decisions a learner must make when working with
demographic and socioeconomic data to predict whether an individual is likely to
exceed a specified income threshold. Questions cover exploratory data analysis,
categorical encoding, feature scaling, train/test splitting, baseline models, decision
trees, ensemble methods, support vector machines, logistic regression, GridSearchCV,
cross-validation, precision, recall, F0.5, confusion matrices, class imbalance, feature
importance, and model selection. Several questions require distinguishing between a
model that achieves strong overall accuracy and one that better satisfies the project's
business objective. The bank also emphasizes avoiding data leakage, interpreting
preprocessing correctly, explaining model behavior, and selecting an appropriate final
estimator. The goal is to develop practical reasoning rather than memorization.
CORE DOMAINS TESTED
1.
CharityML donor-classification problem formulation
2.
3.
Exploratory data analysis and preprocessing
4.
5.
1
,Categorical feature encoding
6.
7.
Numerical feature handling and scaling
8.
9.
Train/test splitting and reproducibility
10.
11.
Baseline and supervised learning models
12.
13.
Decision trees and ensemble methods
14.
15.
Support Vector Machines and Logistic Regression
16.
17.
GridSearchCV and hyperparameter optimization
18.
19.
Cross-validation and model selection
20.
21.
Accuracy, precision, recall, F-score, and F0.5
22.
23.
Confusion-matrix interpretation
24.
25.
2
, Class imbalance
26.
27.
Feature importance and model interpretation
28.
29.
Data leakage and pipeline design
30.
31.
Business-oriented model evaluation
32.
33.
Model efficiency and computational tradeoffs
34.
35.
Final-model justification
36.
QUESTIONS 1-100
Q1: What is the primary supervised-learning objective in the
CharityML donor-classification project?
A) Predict the exact annual income of every individual
B) Predict whether an individual belongs to the higher-income class based on
available features
C) Cluster individuals into demographic groups without labels
D) Reduce the dimensionality of the census dataset
Rationale: The correct answer is B because the task is a binary classification problem
in which the model predicts membership in the target income category. A is incorrect
because exact income prediction would be regression. C describes unsupervised
learning rather than supervised classification. D describes dimensionality reduction,
which may be useful during preprocessing but is not the project's primary objective.
3
