[Seminar] Quantum-inspired manifold learning and feedback-based quantum optimization
Quantum-inspired manifold learning and feedback-based quantum optimization - Dr Mohan Sarovar (Sandia National Labs in California, USA)
In the first act, I will present a new classical algorithm for an important task in machine learning and data science, that of learning the geometric structure of a large dataset . The algorithm is inspired by properties of quantum dynamics of free particles on curved spaces and utilizes classic results in the quantum-classical correspondence. In addition, the algorithm reveals interesting connections between discretization imposed by data sampling and the concept of quantization.
In the second act, I will introduce a new feedback-based strategy for quantum optimization . There is compelling evidence that quantum computers can deliver approximate solutions to discrete optimization problems that are superior to those provided by classical approximation algorithms. In our work, we show that the results of qubit measurements can be used to constructively assign values to variational quantum circuit parameters. Importantly, this measurement-based feedback enables approximate solutions to the combinatorial optimization problem without the need for any classical optimization effort, as would be required for the quantum approximate optimization algorithm (QAOA).
 A. Kumar, M. Sarovar. Manuscript in preparation.
 A. Magann, K. Rudinger, M. Grace, M. Sarovar. https://arxiv.org/abs/2103.08619.
Biography of the speaker:
Mohan Sarovar has been a staff member at Sandia National Laboratories (CA) since 2011. Prior to this he was a post-doctoral scholar at the University of California, Berkeley (2006–2011), and earned his Ph.D. in physics from the University of Queensland, Australia (2003–2006). His research is focused on several areas in quantum information science, including quantum computing, quantum simulation, and quantum communication, and he has contributed to over 60 publications on these topics since 2004 (see https://sites.google.com/site/sarovar for more details). He is currently the PI for several projects studying near-term quantum computing, quantum algorithms for data science, and classical software support for quantum computers.
Join Zoom meeting
Please be sure to register before joining the meeting.
*Attendees from Universities or public/commercial institutions are preferred while anonymous individual users will not be preferred