About the project
This project will incorporate specific choices of configurations of quantum state preparation gates that use integrability to control the search spaces to be explored.
Machine learning methods find efficient solutions in parametric search spaces. In quantum computing applications, the exponentially large representation spaces make generic search intractable.
The advent of noisy intermediate scale (NISQ) quantum computers has prompted the need for applications that can utilise the qualitatively different properties that such devices offer. The algorithmic passage through computational states is engineered by suitably chosen quantum gates and measurements, and the choice of configurations by machine learning methods is viewed as a promising research direction, with the exponentially large Hilbert space offering flexibility for diverse applications. However, finding the appropriate parameterised configuration in this large space can be an obstacle for gradient based machine learning methods.
This research project will investigate a choice of parameterisations for a layered architecture for taming the search space of quantum machine learning models. One layer will be characterised by Yang-Baxter gates that are derived from the study of integrable quantum systems, while the other will be subspace-restricted generic circuits that typically span all of the available phase space. Example applications to be explored in this project will include variational optimisation problems that are viewed as likely first successes for quantum computing methods.
You will have the opportunity to grow your understanding of this rapidly developing subject.
You will also work in a vibrant group environment and be able to interact closely with a community of research students engaged with applications of quantum technologies.