Special AEP Seminar: Oleksandr Kyriienko (Sheffield)

Physics-informed and Physics-based Quantum Machine Learning

In the talk I will describe recent advances in quantum machine learning and discuss possible benefits of physics-based ML pipelines. I will introduce different ways for embedding physics into quantum machine learning (QML) models. Specifically, concentrating on tasks with symmetries, I will show how QML can naturally learn powerful protocols for analyzing data and learning from differential equation solvers. First, I will introduce approaches based on differential constraints and present an analysis for embeddings motivated by physical processes. Second, I will show that embedding symmetries into quantum machine learning models can help discovering protocols for solving the forrelation problem with excellent generalization. Finally, I will present recent results on polaritonic machine learning for graph-based problems, highlighting the potential of feature selection with nonlinear photonic systems.

Bio: Oleksandr Kyriienko is a Ukrainian theoretical physicist working in the areas of quantum computing and quantum optics. He has finished M.Sc. studies at the National University of Kyiv, and in 2010 went for a Ph.D. program at the University of Iceland. Oleksandr completed the Ph.D. in quantum optics in 2014, and from 2014 to 2017 has worked as a postdoctoral researcher at the Niels Bohr Institute in Copenhagen, Denmark. During this time, he concentrated on developing quantum simulation approaches. In 2017 he received a two-year fellowship to conduct an independent research program at NORDITA (Nordic Institute for Theoretical Physics) in Stockholm, Sweden. At that point Kyriienko started developing quantum algorithms for emerging near-term quantum computers. He moved to the University of Exeter as a lecturer (assistant professor) at the end of 2019 and started building a research group working on different subjects of quantum technology. Concentrating on the field of quantum scientific computing and machine learning, together with industrial collaborators he has proposed various protocols for solving PDEs and generative modelling. As of 2025 he is a Chair in Quantum Technologies at the University of Sheffield and Director of the Sheffield Quantum Centre. Oleksandr continues to lead research in quantum SciML and simulation.