Projects
Customer Segmentation: Using K-Means Clustering & Decision Trees to Profile Customer Behavior
Analyzed customer shopping patterns and delivered marketing strategies, retention ideas, and inventory recommendations. Addressed dimensionality using the PCA technique, created RFM profiles, and segmented customers. Using a decision tree model, I identified 6 distinct shopping personas.
Do Lyrics Matter in Songs? Leveraging XGBoost, LSTM, and Lyric Semantics to Predict Skip Behavior
Using personal Spotify data, I investigated whether song lyrics can predict skip behavior. The project involved web scraping, NLP, and correlation analysis. I also used three ML models (Logistic Regression, XGBoost, and LSTM) for predictive modeling. Presented recommendations for streaming platforms, record labels, and data teams on how to improve listening experience.
NBA Tempo & Injuries: Analyzing the Effects on Performance and Winnings
Examined the effects that tempo and pace have on winning and performance in the NBA. Created multiple visualizations in R that looked at avg speed, pace, possession and distance per season. Used correlation plots to view relationship between different game statistics. Developed guidance around managing injuries, and minimizing the cost for the league.