Credit Card Fraud Analytics & Detection

Project Overview

This project analyzes real-world credit card transaction data to identify fraudulent activities using Python, SQL, Power BI, and Machine Learning.

The project includes:

• Exploratory Data Analysis (EDA)
• Fraud trend analysis
• Customer segmentation
• SQL analytics
• Interactive Power BI dashboard
• Fraud prediction model using Logistic Regression

---

Dataset

• Source: Kaggle
• Rows: 786K+ transactions
• Real-world financial transaction dataset

---

Technologies Used

• Python
• Pandas
• NumPy
• SQL (SQLite)
• Power BI
• Scikit-learn
• Matplotlib
• Seaborn

---

Features Engineered

• Risk Score
• Customer Segmentation
• Transaction Hour
• Day of Week
• Monthly Analysis

---

Dashboard Features

• Fraud Rate KPI
• Fraud Transactions KPI
• Top High-Risk Merchants
• Fraud Transactions by Hour
• Customer Segmentation
• Interactive Filters

---

Machine Learning

Model Used:

• Logistic Regression

Evaluation Metrics:

• Accuracy Score
• Confusion Matrix
• Classification Report

---

Project Structure

credit-card-analytics/ │ ├── cleaned_transactions.csv ├── transactions.db ├── eda.ipynb ├── queries.sql ├── credit_card_fraud_dashboard.pbix └── README.md

---

Dataset & Dashboard Files

Large project files are hosted on Google Drive. https://drive.google.com/drive/folders/1ziKw3KdstokbMlEiGJBTTyJcF-2FS5M1

Author

Akhil Agarwal B.Tech AI & Data Science