DATA 6999: MBA Capstone Project Presentation

Illya Mowerman, Ph.D.

2025-05-20

Welcome to DATA 6999 - MBA Capstone

Objective: Create a project you’re proud to showcase to future employers
Course Structure: Flexible, student-driven project selection
Grading: Pass (A), Fail (F), or Incomplete
- A: Complete a feasible project
- F: No work submitted, no withdrawal
- Incomplete: Started but not finished
Instructor Support: Contact via Slack or email (24-hour response time)

Anything You’re Proud Of: Design a project that reflects your skills and interests
Examples:
- Build a website
- Develop a Python pipeline (e.g., data processing, analytics)
- Create a mobile or web app
- Conduct a traditional analytical study
Key Requirement: Project must be feasible within the course timeframe

Focus: Answer a research question using data analysis
Steps:
1. Define a research question
2. Source or acquire relevant data
3. Perform Exploratory Data Analysis (EDA)
4. Clean and prepare data
5. Build and evaluate a model
6. Draw conclusions and interpret results
Example Research Question: “Why do employees leave, and what retention strategies are effective?”

What is it?: A specific question you aim to answer through data analysis
Example: Using HR data to explore employee turnover
- Question: “What factors drive employee attrition?”
- Goal: Uncover insights, not just predictions
Data Sources:
- Find your own dataset (e.g., Kaggle)
- Avoid replicating existing Kaggle analyses; ask a unique question

Exploratory Data Analysis (EDA):
- Visualize data (e.g., histograms, scatter plots)
- Identify patterns, outliers, or trends
Data Cleaning: Handle missing values, inconsistencies
Modeling:
- Choose appropriate algorithms
- Evaluate with metrics (e.g., confusion matrix, ROC curves, sensitivity, specificity)
Conclusions: Interpret model results to answer your research question

Purpose: Understand prior work in your chosen field
What to Include:
- Research on similar problems
- Algorithms or methods used by others
Scope: Not as extensive as an academic paper, but sufficient to contextualize your work
Tip: Cite relevant studies to justify your approach

Structure of the Final Report:
- Abstract: Summarize your project
- Introduction: State the problem and research question
- Literature Review: Summarize prior work
- Data Description: Include data dictionary, source, and fields
- EDA: Share visualizations and findings
- Modeling: Explain model, performance metrics (e.g., ROC, confusion matrix)
- Discussion: Interpret results and implications
Focus: Clear communication of process and insights

Creative Freedom: Build something unique!
Examples:
- Website: Showcase a portfolio or business concept
- Python Pipeline: Automate tasks (e.g., portfolio balancing)
- App Development: Create a functional prototype
Approval Criteria: Feasibility within the course timeframe
Tip: Align with your career goals to maximize impact