A more detailed write up of each project along with my resume can be found on my personal website: https://www.notion.so/welcome-to-my-data-science-portfolio-2075d2b3ab7880fea009eaf700408b20
Goal: Identify distinct Spotify user segments based on demographics, listening behavior, and subscription preferences.
Methods: KModes clustering (categorical data), silhouette & elbow methods, association rule mining (support, confidence, lift).
Tools: Python, pandas, kmodes, mlxtend, matplotlib, seaborn.
π Repository: https://github.com/stjsmith8/user-segmentation
Goal: Predict customer churn and identify behavioral drivers of attrition to support proactive retention strategies.
Methods: Logistic Regression, Random Forest, XGBoost, SMOTE, feature importance, ROC/AUC analysis.
Tools: Python, scikit-learn, XGBoost, pandas, matplotlib, seaborn.
π Repository: https://github.com/stjsmith8/customer-churn-prediction
- Python, R, SQL, SAS
- Statistical Modeling & Machine Learning
- Time Series Forecasting
- Clustering & Segmentation
- Data Visualization (Tableau, ggplot2)
- Program Evaluation & Applied Research
- LinkedIn: https://linkedin.com/in/YOURNAME
- GitHub: https://github.com/USERNAME