Fuzzy Logic-based Explainable AI
Project Background: According to Price Waterhouse Coopers, the global economic impact of Artificial Intelligence (AI) is a mind boggling $15 Trillion, "making it the biggest commercial opportunity in today's fast changing economy" by bringing about the transformation of productivity resulting in GDP growth. To realize this immense economic potential for the safety and mission critical arena such as Health Care, Transportation, Aerospace, Cybersecurity, and Manufacturing existing vulnerabilities need to be clearly identified, and addressed. The end user of such applications as well as the taxpaying public will need assurances that the fielded systems can be trusted to deliver as tasked. Moreover, recent developments evaluating the trustworthiness of high performing black-box” AI has classified them using the term "Brittle AI". These developments coupled with a growing belief in the need for "Explainable AI" has led major policy makers in the US and Europe to underscore the importance of "Responsible AI". Over the past three decades, Dr. Kelly Cohen’s research has been focused on developing fuzzy logic-based AI solutions that are explainable and responsible. Figure 1 illustrates the key principles of Responsible AI.
Research Tasks Proposed:
- Understanding the need for responsible AI and main challenges
- Learning basics of Fuzzy Logic AI programming using MATLAB
- Application of Fuzzy Logic AI to an impactful engineering use case
- Cross-discipline collaboration skills: presentation, writing skills;
- Potential for research publications and presentations at the local and national levels.
Research Location: This research project will be completed in the AI Bio Lab at Digital Futures (third floor, Suite #300)
Key Principles of Responsible AI https://www.altexsoft.com/blog/responsible-ai/
Director
Kelly Cohen
Brian H. Rowe Endowed Chair in Aerospace Engineering , CEAS - Aerospace Eng & Eng Mechanics
745 Baldwin Hall
Dr. Cohen is passionate about mentoring graduate students, having graduated 13 PhD and 37 MS students — many of whom have earned national best-thesis and early-career awards — while currently advising a research team of 18 PhD and 2 MS students. His main expertise lies in explainable and certifiable Artificial Intelligence (AI), intelligent systems, UAVs and advanced air mobility, and optimization. He has pioneered genetic fuzzy logic-based machine-learning algorithms for control and decision-making in autonomous collaborative robotics — including the landmark 2016 ALPHA system, whose technology transitioned to industry through the Thales TrUE AI Toolkit — as well as predictive modeling for personalized medicine, cybersecurity, and safety assurance in safety-critical systems. He has secured grants from NSF, NIH, USAF/AFRL, DHS, DOT, ODOT, OFRN, the State of Ohio, and NASA to develop AI and UAV technologies. His scholarly record includes more than 850 outputs: 76 peer-reviewed journal articles, 118 book chapters, 4 co-edited books, nearly 300 conference papers, and numerous invited seminars, keynotes, and patents.