This portfolio presents selected data-science and analytics projects involving machine learning, exploratory data analysis, predictive modeling, data visualization, and responsible use of data.
My professional background includes enterprise data architecture, database development, systems engineering, and technical program leadership. These projects demonstrate how I apply Python, statistical analysis, machine learning, and visualization to practical questions in education, public policy, housing, and related domains.
This portfolio currently includes six completed projects and three projects that are still being developed or finalized. Project statuses will be updated as additional analysis and documentation are completed.
| Project | Status |
|---|---|
| University Dropout Prediction | Complete |
| Childcare Costs Across the United States | Complete |
| New York City Airbnb Listing Analysis | Complete |
| Netflix Viewership Analysis | Complete |
| White House Attendance Analysis | Complete |
| TSA Complaint Analysis | In Progress |
| Hays County Real Estate Analysis | In Progress |
| Startup Business Outcome Prediction | In Progress |
| EV Purchase Prediction | Complete |
Machine-learning classification project that compares Logistic Regression, Random Forest, and XGBoost for identifying students at risk of withdrawing. The project emphasizes recall, interpretability, fairness evaluation, and responsible institutional use.
Technologies: Python, pandas, scikit-learn, XGBoost, Matplotlib, Jupyter Notebook
View the University Dropout Prediction project
Exploratory analysis of the National Database of Childcare Prices examining weekly childcare prices across states, regions, age groups, and care settings. The repository includes a notebook, dashboard, presentations, and written report.
Technologies: Python, pandas, NumPy, Matplotlib, Seaborn, Jupyter Notebook
View the Childcare Costs Across the United States project
Exploratory analysis of 2019 New York City Airbnb data examining price distributions, room types, review activity, and neighborhood differences.
Technologies: Python, pandas, Matplotlib, Seaborn, Jupyter Notebook
View the New York City Airbnb Listing Analysis project
Exploratory analysis of Netflix Top 10 data examining global viewing trends, sustained chart presence, content categories, country-level coverage, and titles with broad geographic visibility.
Technologies: Python, pandas, Matplotlib, openpyxl, Jupyter Notebook
View the Netflix Viewership Analysis project
Descriptive analysis of White House visitor records from selected months in 2022 and 2023. The project examines recorded activity by month, day of the week, arrival hour, named visitee, and meeting location while emphasizing responsible interpretation and data limitations.
Technologies: Python, pandas, Matplotlib, Jupyter Notebook
View the White House Attendance Analysis project
Technologies: Python, pandas, Matplotlib, Jupyter Notebook
View the TSA Complaint Analysis project
A time-series forecasting project that combines county-level housing inventory data with mortgage-rate information to predict Hays County’s median days on market one month in advance. The analysis compares regression and machine-learning models with simple historical baselines and translates the results into practical guidance for real estate professionals.
Technologies: pandas, numpy, matplotlib, seaborn
View the Hays County Real Estate Analysis project
A machine-learning classification project examining whether startup characteristics available at a defined point in time can predict future business outcomes. The analysis addresses class imbalance, compares baseline and machine-learning models, and evaluates which company and funding characteristics contribute most to the predictions.
Status: Complete
Technologies: Python, pandas, scikit-learn, Matplotlib, Seaborn, Jupyter Notebook
View the Startup Business Outcome Prediction project
Machine-learning classification project predicting electric vehicle purchase interest using demographic, financial, transportation, charging-access, and attitudinal data. The analysis compares logistic regression, random forest, and histogram gradient boosting models, with the final model achieving a validation ROC AUC of 0.9414 and a Kaggle public leaderboard score of 0.94118.
Technologies: Python, pandas, NumPy, scikit-learn, Matplotlib, Jupyter Notebook
View the EV Purchase Prediction project
I am an enterprise data architect and technical leader with extensive experience designing data platforms, integrating complex information systems, and guiding technology initiatives from planning through implementation. My interests include data architecture, predictive analytics, machine learning, data engineering, and the responsible application of analytical models.