OPEN SPACE,
OPEN MIND

[Press Release] Constructing a Data Representation Framework Based on Noncommutative Quantum Dynamics: Systematically Evaluating the Influence of Noncommutativity and Orbits on Representation Structure

Research

[Key points of the research results]
Focusing on quantum dynamics, we constructed a data representation framework for quantum machine learning that utilizes its orbital properties and non-commutativity.
We investigated the influence of non-commutativity and orbitals on the representation, factor by factor, and systematically evaluated the structure and complexity of the representation based on spectral analysis and effective dimensions.
The insights gained will deepen our understanding of data representation in quantum machine learning, contribute to the organization of design guidelines, and demonstrate the potential for application to new quantum machine learning methods.

【Overview】
Associate Professor Hirokazu Shimauchi of Future University Hakodate has constructed a framework for data representation based on orbits in quantum dynamics and systematically investigated the influence of non-commutativity and orbits on the representation structure. Furthermore, he characterized the structure and complexity of the representation through spectral analysis and evaluation based on effective dimension.
The results of this research are, March 27, 2026 It was published in IEEE ACCESS, vol. 14.