📊 Customer Churn Prediction using PyTorch & Logistic Regression

📌 Project Overview

Customer churn is a critical business problem for telecom companies, as retaining existing customers is more cost-effective than acquiring new ones. This project focuses on predicting whether a customer will churn using historical telecom customer data.

Two models were built and compared:

• Logistic Regression (baseline ML model)
• Neural Network using PyTorch

The goal is to evaluate model performance and understand key factors influencing churn.

---

🧠 Problem Statement

Predict whether a customer will churn (Yes/No) based on demographic details, account information, and service usage patterns.

---

📂 Project Structure

customer-churn-pytorch/
│
├── data/
│   ├── raw_data.csv
│   ├── processed_data.csv
│
├── notebooks/
│   ├── 01_eda.ipynb
│   ├── 02_preprocessing.ipynb
│
├── src/
│   ├── linear_regression.py
│   ├── NeuralNetwork.py
│
├── results/
│   ├── confusion_matrix.png
│   ├── metrics.txt
│
├── requirements.txt
├── README.md
└── main.py

---

🔍 Exploratory Data Analysis (EDA)

• Analyzed churn distribution and class imbalance
• Identified numerical vs categorical features
• Observed higher churn in month-to-month contracts and short-tenure customers
• Checked missing values and data consistency

---

🛠 Data Preprocessing

• Converted target variable (Churn) to binary (0/1)
• Handled missing values
• One-hot encoded categorical variables
• Scaled numerical features using StandardScaler
• Addressed class imbalance using class weights

---

🤖 Models Used

Logistic Regression

• Baseline classification model
• Used class weighting to handle imbalance

Neural Network (PyTorch)

• Fully connected feedforward neural network
• Binary classification using BCEWithLogitsLoss

---

📈 Model Evaluation

Metrics used:

• Accuracy
• Precision
• Recall
• F1-score
• Confusion Matrix

Results Summary

Model	Accuracy
Logistic Regression	~75%
PyTorch Neural Network	~77%

---

🚀 How to Run the Project

1. Clone the repository

git clone https://github.com/your-username/Customer-Churn-Prediction.git
cd Customer-Churn-Prediction

2. Install dependencies

pip install -r requirements.txt

3. Run notebooks for EDA and preprocessing

4. Run models

python main.py

---

🧰 Technologies Used

• Python
• Pandas, NumPy
• Scikit-learn
• PyTorch
• Matplotlib, Seaborn

---

📌 Key Learnings

• End-to-end ML workflow
• Handling class imbalance
• Comparing ML and Deep Learning models
• Structuring ML projects for GitHub

---

📬 Author

Mouli Shankar
Aspiring Data Scientist | Machine Learning | Analytics