This repository contains a collection of Python AI projects focused on practical cybersecurity applications. Each project demonstrates how machine learning and AI techniques can be applied to detect, classify, and analyze security threats.
This project uses Logistic Regression to classify cybersecurity events such as normal traffic, phishing, DoS attacks, and malware. The model is trained on numeric features like packet size, failed logins, suspicious port usage, and connection duration.
How it works:
- Each feature is assigned a weight (coefficient) during training.
- The weighted sum of features is transformed into probabilities using a sigmoid function.
- The model predicts the event type with the highest probability.
- Users can input new event data to see predictions and probability scores.
- Implemented in Python using popular libraries like
scikit-learn,pandas, andmatplotlib. - Includes data preprocessing, model training, and evaluation.
- Visualizations to better understand model decisions and feature importance.
- Interactive scripts for testing with custom inputs.