🕵️‍♂️ Fraud Detection in Credit Card Transactions 💳

Welcome to our repository dedicated to detecting fraudulent activities in credit card transactions using advanced machine learning techniques. This project applies various Gradient Boosting algorithms to identify fraud cases from a highly imbalanced dataset.

📚 Dataset

The dataset includes transactions made by credit cards, where each transaction is labeled as fraudulent or legitimate. Key features include:

• Time of transactions
• Transaction Amount
• 28 anonymized features (V1 to V28)
• Class (1: Fraud, 0: No Fraud)
• 🧰 Tools and Libraries

Python 🐍

• Pandas & NumPy for data manipulation
• Matplotlib & Seaborn for data visualization
• Scikit-learn for model building and evaluation
• Imbalanced-learn for handling imbalanced data
• XGBoost, LightGBM, CatBoost, and other boosting algorithms

🔍 Exploratory Data Analysis

• Analysis of transaction amounts and time distribution
• Visualization of fraud vs. no fraud transactions
• Correlation analysis using heatmaps

📉 Visualizations

• Transaction Class Distribution: A bar chart showing the distribution of fraudulent and non-fraudulent transactions.
• Time and Amount Distributions: Histograms and scatter plots for transaction time and amount, divided by class.
• Correlation Heatmap: A heatmap to visualize the correlation between different features.

🤖 Model Building and Evaluation

• Feature scaling and data resampling for balanced dataset
• Model training using Gradient Boosting algorithms like XGBoost, LightGBM, CatBoost
• Evaluation using classification report, confusion matrix, and accuracy scores

🚀 How to Run

Clone the repository Install dependencies Run the Jupyter notebooks or Python scripts