Developed an end-to-end Credit Card Fraud Detection system using Artificial Neural Networks (ANN) to identify fraudulent transactions from highly imbalanced financial data. Implemented comprehensive data preprocessing, feature scaling, class imbalance handling, and hyperparameter tuning to improve fraud detection performance. Built an interactive Streamlit web application that allows users to upload transaction datasets or manually enter transaction details to receive real-time fraud predictions. The model was evaluated using precision, recall, F1-score, ROC-AUC, and Precision-Recall AUC, with threshold tuning to reduce false positives while maintaining high fraud detection capability.

Tech Stack:

Python
TensorFlow / Keras
Scikit-learn
Pandas
NumPy
Streamlit
Plotly
Joblib

Key Features:

ANN-based binary classification model
Data preprocessing and feature scaling using StandardScaler
Class imbalance handling using class weights
Hyperparameter tuning and Early Stopping
Threshold optimization for improved fraud detection
Interactive Streamlit dashboard
CSV upload and manual transaction prediction
Performance visualization using confusion matrix, ROC curve, and Precision-Recall curve