Color Teams for Machine Learning Development

TL;DR

This paper introduces a novel 'Color Teams' framework for machine learning development, inspired by cybersecurity practices, to enhance robustness and shared knowledge among team members.

Contribution

It proposes a new team structuring approach with color-coded roles, including new team types, to improve collaboration and security in machine learning projects.

Findings

Color Teams improve collaboration and robustness.

New team roles enhance cybersecurity in ML development.

Framework facilitates shared knowledge and responsibility.

Abstract

Machine learning and software development share processes and methodologies for reliably delivering products to customers. This work proposes the use of a new teaming construct for forming machine learning teams for better combatting adversarial attackers. In cybersecurity, infrastructure uses these teams to protect their systems by using system builders and programmers to also offer more robustness to their platforms. Color teams provide clear responsibility to the individuals on each team for which part of the baseline (Yellow), attack (Red), and defense (Blue) breakout of the pipeline. Combining colors leads to additional knowledge shared across the team and more robust models built during development. The responsibilities of the new teams Orange, Green, and Purple will be outlined during this paper along with an overview of the necessary resources for these teams to be successful.