Skip to content

Repository files navigation

Application of AI for InfoSec: Applied ML Projects

Four ML projects applying classic and deep learning techniques to security and text-classification problems, built while working through Hack The Box's AI Red Teamer path (Module 2: Applications of AI in InfoSec).

Projects

Project Technique Dataset
Spam Classification Naive Bayes + CountVectorizer SMS Spam Collection
Network Anomaly Detection Random Forest (multi-class) NSL-KDD
Malware Classification CNN (ResNet-50, transfer learning) Malimg
Sentiment Analysis Naive Bayes + TfidfVectorizer IMDB Reviews

Each folder contains a notebook and a short write-up covering approach and results.

Setup

conda create -n ai-infosec python=3.11
conda activate ai-infosec
pip install -r requirements.txt

Or with pip directly:

pip install -r requirements.txt

Notes

The first three projects each use a meaningfully different ML approach — text classification with Naive Bayes, tabular multi-class classification with Random Forest, and image-based classification with a fine-tuned ResNet-50. The sentiment analysis project is a skills-assessment exercise using the same pipeline as the spam classifier, applied independently to a new dataset.

Credit

Based on Module 2 : Applications of AI in InfoSec of the AI Red Teamer Job Role Path by Hack The Box, developed in collaboration with Google.

Description

About

Hands-on ML for security: spam classification, network anomaly detection, and malware-as-image classification.

Resources

Stars

1 star

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages