📉 Customer Churn Prediction App

A complete end-to-end Machine Learning project that predicts whether a customer is likely to churn, built with Python, Scikit-learn, and deployed as an interactive web application using Streamlit.

---

🚀 Live Demo

Run the app locally by following the setup instructions below.

---

📌 Project Overview

Customer churn is one of the biggest challenges businesses face. Losing a customer is far more expensive than retaining one. This project uses historical customer data to train a Machine Learning model that identifies customers at high risk of churning — enabling businesses to take proactive retention steps.

This project covers the full ML pipeline:

• Data loading & exploration
• Data preprocessing & feature engineering
• Model training & evaluation
• Saving the trained model
• Deploying as an interactive web app

---

🛠️ Tech Stack

Tool	Purpose
Python	Core programming language
Pandas & NumPy	Data manipulation & analysis
Scikit-learn	ML model training & evaluation
Streamlit	Web app deployment
Matplotlib / Seaborn	Data visualization
Jupyter Notebook	Exploratory Data Analysis (EDA)

---

📁 Project Structure

Churn-Prediction-App/
│
├── app.py                    # Streamlit web application
├── notebook.ipynb            # EDA, preprocessing & model training
├── model.pkl                 # Trained ML model (serialized)
├── scaler.pkl                # Fitted scaler for input normalization
├── customer_churn_data.csv   # Dataset used for training
└── README.md                 # Project documentation

---

⚙️ How to Run Locally

1. Clone the repository

git clone https://github.com/irlhasnain/Churn-Prediction-App.git
cd Churn-Prediction-App

2. Install dependencies

pip install -r requirements.txt

If you don't have a requirements.txt, install manually:

pip install streamlit pandas numpy scikit-learn matplotlib seaborn

3. Run the Streamlit app

streamlit run app.py

4. Open your browser at http://localhost:8501

---

🧠 Model Details

• Algorithm: Random Forest Classifier (/ Logistic Regression — update as applicable)
• Target Variable: Churn (1 = Churned, 0 = Retained)
• Key Features Used: Contract type, tenure, monthly charges, payment method, internet service, etc.
• Preprocessing: Standard scaling applied via scaler.pkl

---

📊 Model Performance

Best Model: SVM (C=0.01, Linear Kernel) — selected via GridSearchCV (5-fold cross-validation)

Metric	Score
Accuracy	84%
Precision (Churn Class)	84%
Recall (Churn Class)	100%
F1 Score (Churn Class)	91%
Weighted Avg F1	77%

⚠️ Note: The model predicts churn (class 1) with high recall (100%) — meaning it catches almost every customer who will churn. This is intentional, as missing a churner is more costly for a business than a false alarm.

---

💡 Key Learnings

• Built a real-world ML pipeline from raw data to deployed application
• Handled class imbalance and feature encoding for categorical variables
• Serialized trained model using pickle for deployment
• Created an intuitive UI with Streamlit for non-technical users

---

🙋‍♂️ Author

Hasnain Khan
Data Science Aspirant | Python • ML • SQL • Power BI
📍 Bhopal, Madhya Pradesh, India

[ [

---

📄 License

This project is open source and available under the MIT License.