Predictive Customer Churn Analysis for Telecom Industry
www.kaggle.com/code/michaelenefiok/customer-churn-prediction-with-ml
Summary
1. Analyzed 7,000+ customer records to model churn behavior and reduce revenue loss for a telecom provider. 2. Engineered features, handled class imbalance, and trained Random Forest and Logistic Regression models, achieving 86% accuracy and 0.82 ROC-AUC. 3. Visualized insights in matplotlib/seaborn to communicate drivers of churn contract type, tenure, and monthly charges to non-technical stakeholders. 4. Deployed findings in a Kaggle notebook with interactive charts.