CQuIC Seminars
Quantum-inspired geometric data analysis
Presented by Mohan Sarovar (Sandia)
I will begin with brief advertisements for three recent papers from our group, Refs. [1]-[3].
Then I will focus the rest of the seminar on the topic of learning from large datasets. Learning the structure and organization of data is a ubiquitous task in modern science and is often the first step in many machine learning algorithms. Recently, we proposed that much like how quantum mechanics models nature at fine scales, the fine-scaled resolution of the organization and structure of data is also best characterized using quantum mechanical processes [4].
This proposal is realized through the development of a learning algorithm that proceeds by simulation of quantum dynamics on a graph-embedding of data, and the establishment of a quantum-classical correspondence between data-driven dynamics on data and geodesic flow on the underlying data manifold. Conceptually, these results connect the notion of discretization imposed by data sampling to the concept of quantization in physics.
I will explain this quantum data analysis framework and present some of the results we have obtained using it on model and real-world datasets. Finally, I will briefly discuss ongoing research attempting to utilize this framework to develop quantum algorithms for data analysis.
[1] Surrogate-based optimization for variational quantum algorithms. Ryan Shaffer, Lucas Kocia, Mohan Sarovar. arXiv:2204.05451 (2022).
[2] Establishing trust in quantum computations. Timothy Proctor, Stefan Seritan, Erik Nielsen, Kenneth Rudinger, Kevin Young, Robin Blume-Kohout, Mohan Sarovar. arXiv:2204.07568 (2022).
[3] Digital adiabatic state preparation error scales better than you might expect. Lucas Kocia, Fernando A. Calderon-Vargas, Matthew D. Grace, Alicia B. Magann, James B. Larsen, Andrew D. Baczewski, Mohan Sarovar. arXiv:2209.06242 (2022).
[4] Manifold learning via quantum dynamics. Akshat Kumar, Mohan Sarovar. arXiv:2112.11161 (2021).
3:30 pm, Thursday, December 8, 2022
PAIS-2540, PAIS
Individuals with disabilities who need an auxiliary aid or service to attend or participate in P&A events should contact the Physics Department (phone: 505-277-2616, email: physics@unm.edu) well in advance to ensure your needs are accomodated. Event handouts can be provided in alternative accessible formats upon request. Please contact the Physics front office if you need written information in an alternative format.
A schedule of talks within the Department of Physics and Astronomy is available on the P&A web site at http://physics.unm.edu/pandaweb/events/index.php