Is machine learning a solution to the problem of finding attackers in haystacks of data or is it merely another buzzword? In this workshop, participants will learn how machine learning can solve cybersecurity problems by constructing their own classification models. We will follow the complete model building workflow, starting with data cleaning, then continuing to feature engineering, construction of multiple models, and evaluation of models. We will use Python's scikit-learn package to construct models with real world security data.

Workshop materials in ZIP are available.

Copyright and Licensing

© 2018 James Walden. All materials on this page are available under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 Unported License. Some slides are based on the POSSE 2014-5 workshop slides, which are available under the same license.