💳 decision_tree_svm_creditcard_fraud

📖 Overview

Compare Decision Tree and SVM models for detecting credit card fraud. Includes data preprocessing, model training, evaluation, and visualizations.

📊 Dataset

Credit card transactions labeled as fraudulent or non-fraudulent. Features are anonymized via PCA.

Source: Credit Card Fraud Detection Dataset

🎓 Credit

Inspired by IBM Machine Learning Professional Certificate on Coursera:
IBM Developer Skills Network

🛠 Steps

1. Load dataset & visualize class distribution.
2. Feature correlation analysis.
3. Standardize & normalize features.
4. Train-test split & sample weighting.
5. Train models: Decision Tree (max_depth=4) & Linear SVM.
6. Evaluate with ROC-AUC, ROC curve, and Precision-Recall curve.

📈 Results

Model	ROC-AUC
Decision Tree	0.939
SVM	0.986

🚀 How to Run

git clone https://github.com/<your-username>/decision_tree_svm_creditcard_fraud.git
cd decision_tree_svm_creditcard_fraud
pip install -r requirements.txt
# Open notebook or run script