Built a logistic regression model with SMOTE sampling, improving prediction fidelity by 87%.
Visualized churn insights using violin plots and swarm graphs to enhance interpretability.