Credit-Card-Fraud-Detection

This project demonstrates a machine learning model to detect fraudulent credit card transactions using a logistic regression model. The dataset is heavily imbalanced, so techniques like SMOTE (Synthetic Minority Over-sampling Technique) are used to oversample the minority (fraud) class. The project includes data preprocessing, scaling, oversampling, training, evaluation, and visualization of the model's performance.

Table of Contents

• Project Overview
• Technologies Used
• Dataset
• Installation
• Usage
• Model Evaluation
• Contributing
• License

Project Overview

The goal of this project is to build a machine learning model that can detect fraudulent credit card transactions. The project involves:

• Loading and preprocessing the data.
• Addressing the class imbalance using SMOTE.
• Training a logistic regression model to classify transactions as fraudulent or non-fraudulent.
• Evaluating the model using accuracy, precision, recall, and F1-score metrics.

Technologies Used

• Python: Version 3.x
• Pandas: For data manipulation and analysis
• Scikit-learn: For machine learning models and evaluation metrics
• Imbalanced-learn: For oversampling techniques (SMOTE)
• Matplotlib: For data visualization

Dataset

The dataset used in this project is a credit card transaction dataset (creditcard.csv) which contains transactions labeled as either fraudulent (Class=1) or non-fraudulent (Class=0). It is highly imbalanced, with only a small percentage of fraudulent transactions.

• Features:
  • Time, Amount, and 28 anonymized features (V1, V2, ..., V28)
  • Target variable: Class (1 for fraud, 0 for non-fraud)

Installation

1. Clone the repository:

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

2. Install the required dependencies:

   pip install -r requirements.txt
   

3. Download the dataset and place it in the root directory of the project. The dataset should be named creditcard.csv.

Usage

After cloning the repository and setting up the environment, follow the steps below to run the project: Preprocess the data: This includes loading the data, scaling the Amount feature, and oversampling using SMOTE. Train the model: The logistic regression model is trained on the resampled data. Evaluate the model: After training, the model is tested, and evaluation metrics such as accuracy, precision, recall, and F1-score are calculated. Visualize the results: A bar chart is generated to compare precision, recall, and F1-scores between the two classes.

Model Evaluation

The model is evaluated using the following metrics:

• Accuracy: Overall performance of the model.
• Precision: The ratio of true positives to the sum of true and false positives.
• Recall: The ratio of true positives to the sum of true positives and false negatives.
• F1-Score: The harmonic mean of precision and recall.
• A bar chart is generated to visualize the comparison of precision, recall, and F1-scores between non-fraudulent and fraudulent classes.

Contributing

Contributions are welcome! If you'd like to contribute to this project, please follow these steps:

1. Fork the repository.
2. Create a new branch (git checkout -b feature-branch).
3. Make your changes and commit them (git commit -m 'Add new feature').
4. Push to the branch (git push origin feature-branch).
5. Open a pull request.

License

This project is licensed under the MIT License - see the LICENSE file for details.