Skip to content

model_summary.md

Latest commit

 

History

History
59 lines (44 loc) · 1.31 KB

model_summary.md

File metadata and controls

59 lines (44 loc) · 1.31 KB

Model Summary – Customer Churn Prediction

Author: Anusha Durgam

📌 Problem

Classify whether a telecom customer will churn based on service usage and demographic data.

🧾 Dataset

  • 7,000+ records, 20+ features
  • Target: Churn (Yes/No)

🔍 Models Evaluated

  • Logistic Regression
  • Random Forest
  • XGBoost
  • SVM ✅ (Best Performance)

⚙️ Preprocessing

  • One-hot encoding for categorical features
  • StandardScaler for numerical
  • Tenure Group bucketing for feature enhancement

🎯 Evaluation Metrics

  • Accuracy
  • F1-score
  • Precision/Recall
  • ROC-AUC
  • Confusion Matrix

✅ Final Model

  • SVM
  • High accuracy & generalization
  • Deployed via Streamlit

📊 Business Dashboard

  • Tool: Power BI
  • File: churm analysis.pbix
  • Insights:
    • Churn breakdown by contract type
    • MonthlyCharges vs Tenure
    • Customer segmentation filters

🚀 Deployment

  • Model served using Streamlit
  • UI includes customer input + result
  • Dashboard supports decision-making for stakeholders

🔚 Summary

This project delivers a complete ML pipeline with real-time prediction and executive-level reporting. The integration of Streamlit and Power BI enables both data science exploration and stakeholder engagement.