Organized by the University of Chicago’s Eric and Wendy Schmidt AI in Science Postdoctoral Fellowship Program.
Yuehaw Khoo works on developing computational and data-driven techniques for problems in biological and physical sciences. In particular, he develops methods for many-body physics, protein structure determination from NMR spectroscopy and Cryo-EM. He is interested in techniques based on (1) convex and non-convex optimization, (2) neural-network and tensor-network methods.
Event: AI+Science Schmidt Fellows Speaker Series: Yuehaw Khoo
Event Date: October 3, 2023
Event Time: 4:30pm – 6:00pm
4:30pm – 5:15pm: Presentation: Randomized tensor-network algorithms for random data in high-dimensions
5:15pm – 5:30pm: Q&A
5:30pm – 6:00pm: Reception
Tensor-network ansatz has long been employed to solve the high-dimensional Schrödinger equation, demonstrating linear complexity scaling with respect to dimensionality. Recently, this ansatz has found applications in various machine learning scenarios, including supervised learning and generative modeling, where the data originates from a random process. In this talk, we present a new perspective on randomized linear algebra, showcasing its usage in estimating a density as a tensor-network from i.i.d. samples of a distribution, without the curse of dimensionality, and without the use of optimization techniques. Moreover, we illustrate how this concept can combine the strengths of particle and tensor-network methods for solving high-dimensional PDEs, resulting in enhanced flexibility for both approaches.