Michele Caprio

Michele Caprio is a Lecturer (Assistant Professor) in Machine Learning and Artificial Intelligence at The University of Manchester. He holds a PhD in Statistics from Duke University (2022) and completed a postdoctoral fellowship at the University of Pennsylvania’s Department of Computer and Information Science. His research focuses on probabilistic machine learning, particularly imprecise probabilistic techniques for uncertainty quantification. He is actively accepting PhD students for projects such as 'Multimodal Credal Learning Theory' and is a member of the London Mathematical Society.

Education:

• PhD in Statistics, Duke University (2022)
• Postdoctoral Researcher, University of Pennsylvania (Computer and Information Science)

Research Interests: Dr. Caprio’s work bridges theoretical foundations and practical applications in machine learning, emphasizing uncertainty quantification through imprecise probabilities. Key areas include conformal prediction, distributionally robust optimization, and the ergodicity of Markov processes. His methods address challenges in ambiguous ground truth, distribution shifts, and epistemic uncertainty in AI systems.

Scientific Awards:

• IJAR Young Researcher Award (2023)
• IMS New Researchers Travel Award (2023)
• IMS Hannan Travel Award (2022)
• Aleane Webb Dissertation Research Fellowship (2021)

Advising & Projects: As PI of the MCAIF: Centre for AI Fundamentals project, he collaborates with researchers and students on AI fundamentals. His advisees include over 15 PhD candidates focusing on topics like distribution shift recovery, imprecise neural networks, and continual learning.

Labs/Teams: Central to his work is the MCAIF initiative, a multidisciplinary team advancing theoretical and applied AI research through collaborative projects.