fraudproject

metrics and fraud detection

Project Title

Fraud Detection Using Machine Learning

📌 Problem Statement

Online payment systems face significant losses due to fraudulent transactions. The goal of this project is to build a supervised machine learning model that can identify fraudulent transactions while minimizing false positives.

📊 Dataset

• Source: Synthetic Data (generated for learning purposes)
• Size: 21497 rows, 23 features
• Target variable: is_fraud (1 = Fraud, 0 = Legit)

🛠 Tools & Technologies

• Python
• Pandas, NumPy
• Scikit-learn
• SQL (for data extraction)
• Matplotlib / Seaborn

⚙️ Approach

1. Data cleaning and preprocessing
2. Exploratory Data Analysis (EDA)
3. Feature engineering
4. Model training (Logistic Regression, XGBoost)
5. Model evaluation using Precision, Recall, ROC-AUC

📈 Key Results

• Precision:
• Recall:
• ROC-AUC:

The model effectively balances fraud detection accuracy and business risk.

▶️ How to Run