---
title: Telco Customer Churn Prediction
emoji: 📞
colorFrom: blue
colorTo: red
sdk: gradio
sdk_version: 4.0.0
app_file: app.py
pinned: false
license: mit
tags:
- machine-learning
- customer-churn
- classification
- business-analytics
- scikit-learn
- telecommunications
datasets:
- telco-customer-churn
metrics:
- accuracy
- precision
- recall
- f1
- roc_auc
pipeline_tag: tabular-classification
---

# 📞 Telco Customer Churn Prediction Model

A machine learning model that predicts customer churn probability for telecommunications companies with **84.2% ROC-AUC** accuracy.

## 🚀 Quick Start

```python
import joblib
import pandas as pd

# Load the model
model = joblib.load('model.joblib')

# Make prediction (requires 45 features)
churn_probability = model.predict_proba(customer_data)[:, 1]
```

## 📊 Model Performance

| Metric | Score |
|--------|-------|
| **ROC-AUC** | **0.8419** |
| **Accuracy** | 0.8055 |
| **Precision** | 0.6572 |
| **Recall** | 0.5588 |
| **F1-Score** | 0.6040 |

## 🎯 Risk Categories

- 🚨 **High Risk (≥80%)**: 49 customers - *Immediate action required*
- ⚠️ **Medium Risk (50-79%)**: 1,512 customers - *Proactive retention*
- 📊 **Low Risk (30-49%)**: 1,110 customers - *Loyalty programs*
- ✅ **Safe (<30%)**: 4,372 customers - *Upselling opportunities*

## 🔧 Features (45 total)

The model uses 45 features including:
- Demographics (Age, Gender, Partner, Dependents)
- Services (Phone, Internet, Security, Streaming)
- Account (Contract, Payment method, Billing)
- Financial (Tenure, Monthly charges, Total charges)

## 💼 Business Impact

- **Customer Retention**: Identify at-risk customers early
- **Revenue Protection**: Prevent churn-related revenue loss
- **Campaign Optimization**: Target retention efforts effectively
- **ROI**: Expected 300-400% return with targeted campaigns

## 🏗️ Model Details

- **Algorithm**: Logistic Regression
- **Training Data**: 5,634 samples
- **Test Data**: 1,409 samples
- **Training Date**: 2025-07-10

## 📋 Usage

See the demo above for interactive predictions!

For batch processing:
```python
# Load model
model = joblib.load('model.joblib')

# Prepare data with all 45 features
results = model.predict_proba(customer_features)[:, 1]

# Categorize risk
def categorize_risk(prob):
    if prob >= 0.8: return "High Risk"
    elif prob >= 0.5: return "Medium Risk"
    elif prob >= 0.3: return "Low Risk"
    else: return "Safe"
```

## 📜 License

MIT License - Free for commercial and academic use.