💳 Credit Card Fraud Detection Project

📖 Project Overview

This project is an end-to-end machine learning model that detects fraudulent credit card transactions. I built this to practice data preprocessing, handling imbalanced data using SMOTE, and comparing the performance of different classification models like Logistic Regression, Random Forest, and XGBoost.

📊 Dataset

The project uses the "Credit Card Fraud Detection" dataset from Kaggle.

• Link: Kaggle Credit Card Fraud Detection Dataset

🚀 Features

• Complete machine learning pipeline for fraud detection
• Handles extreme class imbalance using appropriate evaluation metrics
• Compares multiple algorithms: Logistic Regression, Random Forest, and XGBoost
• Includes preprocessing, model training, and evaluation

📂 Folder Structure

credit-card-fraud-detection/
├── images/               # Visualizations 
├── models/               # Trained models (.pkl files)
├── notebooks/            # Jupyter notebook with full implementation
├── .gitignore
├── README.md            
└── requirements.txt      # Dependencies

---

🛠️ Setup and Installation

1. Clone the repository:

   git clone https://github.com/arshadmurtaza03/credit-card-fraud-detection.git
   cd credit-card-fraud-detection

2. Create and activate a virtual environment:

       # Linux/macOS
   python3 -m venv venv
   source venv/bin/activate
   
       # Windows
   python -m venv venv
   venv\Scripts\activate

3. Install dependencies:

   pip install -r requirements.txt

4. Download the dataset from the Kaggle link above and place it in the data/ directory.

5. Run the Jupyter Notebook

   jupyter notebook notebooks/fraud_detection.ipynb

Usage

All the code and analysis can be found in the Jupyter Notebook inside the notebooks/ directory.

---

Model Evaluation Results

The models were evaluated based on their ability to correctly identify fraudulent transactions (Recall) while maintaining reasonable precision.

Model	Precision (Fraud)	Recall (Fraud)	F1-Score (Fraud)
Logistic Regression	0.058	0.918	0.109
Random Forest	0.871	0.827	0.848
XGBoost	0.728	0.847	0.783

---

How to cite / contact

• Author: Arshad Murtaza
• GitHub: https://github.com/arshadmurtaza03/credit-card-fraud-detection.git
• Email: arshadmurtaza2016@gmail.com