WGU D502 Data Analytics Capstone | BHN1 Task 2: Project Proposal
WGU D502
Data Analytics Capstone
BHN1 Task 2: Project Proposal
Predicting Online Purchase Intention from E-Commerce
Session Behavior Using Machine Learning
Primary analytical method: Random Forest classification
Dataset: UCI Online Shoppers Purchasing Intention Dataset
2026/2027
D502 | BHN1 Task 2 | Page 1
, WGU D502 Data Analytics Capstone | BHN1 Task 2: Project Proposal
Project Alignment Summary
This proposal carries forward the approved D502 capstone topic: predicting whether an e-commerce
browsing session will result in a purchase. The project uses a real, shareable public dataset, a supervised
machine-learning model, a prespecified performance benchmark, and a clearly defined organizational use
case. The same research question, analytical method, benchmark, visualizations, and governance
approach are intended to remain consistent through BHN1 Task 3.
Element Locked project decision
Research question To what extent can a supervised machine-learning
classification model predict whether an online shopping
session will result in a purchase using behavioral and
contextual session characteristics?
Data UCI Online Shoppers Purchasing Intention Dataset; 12,330
sessions; Revenue is the binary outcome.
Primary model Random Forest classifier.
Supporting baseline Logistic Regression for comparative context; it is not the
primary hypothesis-supporting model.
Primary success metric F1 score on an untouched held-out test set.
Prespecified benchmark F1 >= 0.60.
Planned visual communication Class-distribution chart, behavioral-feature comparison,
confusion matrix, and feature-importance chart.
D502 | BHN1 Task 2 | Page 2
WGU D502
Data Analytics Capstone
BHN1 Task 2: Project Proposal
Predicting Online Purchase Intention from E-Commerce
Session Behavior Using Machine Learning
Primary analytical method: Random Forest classification
Dataset: UCI Online Shoppers Purchasing Intention Dataset
2026/2027
D502 | BHN1 Task 2 | Page 1
, WGU D502 Data Analytics Capstone | BHN1 Task 2: Project Proposal
Project Alignment Summary
This proposal carries forward the approved D502 capstone topic: predicting whether an e-commerce
browsing session will result in a purchase. The project uses a real, shareable public dataset, a supervised
machine-learning model, a prespecified performance benchmark, and a clearly defined organizational use
case. The same research question, analytical method, benchmark, visualizations, and governance
approach are intended to remain consistent through BHN1 Task 3.
Element Locked project decision
Research question To what extent can a supervised machine-learning
classification model predict whether an online shopping
session will result in a purchase using behavioral and
contextual session characteristics?
Data UCI Online Shoppers Purchasing Intention Dataset; 12,330
sessions; Revenue is the binary outcome.
Primary model Random Forest classifier.
Supporting baseline Logistic Regression for comparative context; it is not the
primary hypothesis-supporting model.
Primary success metric F1 score on an untouched held-out test set.
Prespecified benchmark F1 >= 0.60.
Planned visual communication Class-distribution chart, behavioral-feature comparison,
confusion matrix, and feature-importance chart.
D502 | BHN1 Task 2 | Page 2
