Can You Rely on Your AI? Applying the AIR Tool to Improve Classifier Performance

Linda Parker Gates is the principle investigator for a long-term research project on AI robustness at the SEI. She leads the Software Acquisition Pathways Initiative in the SEI Software Solutions Division. She specializes in strategic planning, change management, technology transition, and performance excellence, supporting numerous government organizations developing and adopting improvement strategies. She is the chair of the International Association for Strategic Planners Annual Global Conference.

Crisanne Nolan is an Agile transformation engineer in the Continuous Delivery of Capability Directorate at the SEI. She contributes to the SEI transition of Agile research and practices into complex government programs fielding cyber-physical systems through strategic planning, facilitation, and Agile operations. Her research interests include adoption support for rapid capability delivery and human-centered design approaches for business agility.

Mike Konrad is a principal researcher at the SEI. For the past seven years, he has been applying causal discovery and inference to software engineering and cybersecurity questions. Previously, he was on the teams that developed the CMM for Software and CMMI models, publishing several articles and books along the way. He is a Lifetime Member of the IEEE and has a PhD in mathematics (1978).

Nicholas Testa is a senior researcher at the SEI, specializing in AI/ML and causal inference for over two years. Prior to joining the SEI, he was a data scientist at Highmark, Inc., where he applied AI/ML solutions within the Pharmacy Department and occupied a research preceptor position, in which he helped mentor pharmacy students through their analytics rotation and subsequent publication of their results. He holds a dual PhD in integrative and evolutionary biology (2016).

Suzanne (SuZ) Miller is a principal researcher at the SEI working in the Continuous Deployment of Capabilities Directorate. Her current research focuses on synthesizing effective technology transition and management practices from research and industry into effective techniques for use in the governance of programs adopting or contemplating the adoption of Agile or Lean methods.

David Shepard has made a career working in many different areas of the information-technology field. Since 2010, he has worked as a software developer within the SEI Software Solutions Division. Shepard has spent time building networks, administering servers, designing software, writing and debugging software, working on process-improvement initiatives, auditing application security, implementing big-data and machine-learning systems, and assisting digital-forensic investigations. Shepard has worked in corporate, education, and government sectors for large institutions, mid-sized companies, and startups on projects ranging in size from the tens of thousands to the tens of millions of dollars in budget.