Network Security MLOps End-to-End Project скачать в хорошем качестве

Network Security MLOps End-to-End Project

In this video, I walk you through my project on Malicious URL Detection, explaining the architecture, tools, and technologies used to build it. This project aims to detect harmful URLs that can pose risks to users, leveraging the power of machine learning and robust MLOps practices. 🔍 Problem Statement Malicious URLs are links designed by cybercriminals to harm users through phishing emails, social engineering, and other tactics. This project builds a solution to classify URLs as safe or unsafe. 🛠️ Tools and Technologies Used Frontends: Streamlit: For user interaction and single URL predictions. FastAPI: For backend functionality like training and batch predictions. Data Storage and Processing: MongoDB: Storing URL features and labels. pandas: Data manipulation. tldextract: Extracting URL features. Modeling: XGBoost: For training the machine learning model. KNNImputer: Handling missing values during preprocessing. MLOps Components: CI/CD: Implemented with GitHub Actions. Self-hosted Runner: On an AWS EC2 instance. AWS S3: Storing artifacts and models. Docker: Containerization for deployment. Airflow: Automating workflows with DAGs. Deployment: AWS ECR: Storing Docker images. Docker Containers: Running on AWS EC2 instances github : https://github.com/Mayankpratapsingh0...