🔍 Fraud Detection Model & Dashboard

📌 Project Overview

This project tackles the challenge of identifying fraudulent transactions using machine learning. It includes:

• Model training & evaluation in internship.ipynb
• A Streamlit dashboard for real-time predictions 🎛️

---

📊 Dataset

📂 The dataset is hosted on Kaggle:
👉 Kaggle Dataset – Fraud Detection

Due to size constraints, it’s not uploaded to this repository.

---

🧹 Preprocessing Steps

• Cleaned missing values
• Encoded categorical data
• Scaled numerical values using MinMaxScaler
• Split into training and testing sets

---

🤖 Models & Performance

Model	Accuracy
Logistic Regression	99.87%
Decision Tree	99.95%
XGBoost	99.95%

📈 Evaluation metrics used: Accuracy, Precision, Recall, F1-Score

---

🚀 Streamlit App Deployment

The app is deployed using Streamlit and allows real-time fraud prediction. Try it here:
🌐 Live App – Fraud Detection Dashboard

🧰 To run locally:

1. Clone the repo
2. Download dataset from Kaggle and place it in the project root
3. Run:

pip install -r requirements.txt
streamlit run streamlit_app.py