Скачать или смотреть Build a Credit Card Fraud Detection System | Machine Learning Course Intro

🚀 Welcome to the Credit Card Fraud Detection Course!

In this video, I’ll introduce the challenge of detecting fraud in highly imbalanced data and show you what we’ll build together: a production-ready machine learning system that achieves a 91% F1-score on the famous Kaggle credit card fraud dataset.

🔑 What You’ll Learn in This Course:

Explore 284,807 real transactions (only 492 are fraud – 0.17%)

Solve the accuracy paradox in imbalanced datasets

Train 15+ machine learning models (Logistic Regression, Random Forest, XGBoost, Neural Nets, Graph Neural Networks)

Build a real-time dashboard with Streamlit

Deploy an API with FastAPI + Docker

Showcase a portfolio-ready project for employers

📊 By the End:

You’ll master advanced ML techniques (SMOTE, ensembles, anomaly detection)

You’ll understand why accuracy isn’t enough for fraud detection

You’ll have a complete system: data → models → dashboard → API

🛠️ Tools & Setup:

Python 3.8+, Jupyter/VS Code

Dataset: Kaggle Credit Card Fraud Dataset https://www.kaggle.com/datasets/mlg-ulb/cr...

Code: GitHub Repository https://github.com/ysimokat/Bank-Fraud-Det...
#machinelearning #anomalydetection #datascience #deeplearning
👉 Subscribe & hit the bell to follow along with the full series!
👉 Next video: Data Exploration – Understanding 284,807 Transactions