Spotify YouTube
Python SQLite Data Viz

Spotify × YouTube Music Analysis

Cross-platform analysis comparing artist popularity between Spotify and YouTube. Built a SQLite database from two APIs, calculated engagement metrics including views-per-listener ratio and average views per video, and created visualizations for the top 25 artists. Conducted at University of Michigan for SI 206.

Stress Level vs Sleep Hours — 10,000 Records
Python Statistics XGBoost SHAP

Global Coffee & Health Analysis

Statistical analysis of 10,000 records across 20 countries exploring whether coffee consumption impacts stress, sleep, and BMI. Applied chi-squared testing, OLS regression, and XGBoost classification with SHAP feature importance. Key finding: sleep hours, not coffee, was the strongest stress predictor. SI 385 at University of Michigan.