Machine Learning in Cybersecurity: A Guide
September 2019 • Technical Report
Jonathan Spring, Joshua Fallon, April Galyardt, Angela Horneman, Leigh B. Metcalf, Ed Stoner
This report suggests seven key questions that managers and decision makers should ask about machine learning tools to effectively use those tools to solve cybersecurity problems.
Software Engineering Institute
CMU/SEI Report Number
DOI (Digital Object Identifier):10.1184/R1/12363089.v1
This report lists relevant questions that decision makers should ask of machine-learning practitioners before employing machine learning (ML) or artificial intelligence (AI) solutions in the area of cybersecurity. Like any tool, ML tools should be a good fit for the purpose they are intended to achieve. The questions in this report will improve decision makers’ ability to select an appropriate ML tool and make it a good fit to address their cybersecurity topic of interest. In addition, the report outlines the type of information that good answers to the questions should contain. This report covers the following questions:
- What is your topic of interest?
- What information will help you address the topic of interest?
- How do you anticipate that an ML tool will address the topic of interest?
- How will you protect the ML system against attacks in an adversarial, cybersecurity environment?
- How will you find and mitigate unintended outputs and effects?
- Can you evaluate the ML tool adequately, accounting for errors?
- What alternative tools have you considered? What are the advantages and disadvantages of each one?