Tue27Mar20184:00 pmLewis Hall 101
Colloquium: Data Science and Quantum Gravity
Department of Mathematics
University of California — San Diego
Data Science and Quantum Gravity
Data, even “big data”, is finite, and thus discrete. A common goal is to describe them as the outcome of a random process specified by a small number of parameters; doing so at least compresses the data, and at best explicates the process by which they were generated. Some important approaches include low-rank matrix factorization and multi-dimensional scaling, both of which reveal a geometry behind the data. Such interplay between the discrete and the continuous is familiar in theoretical and computational physics, from the definition and regularization of path integrals to numerical methods for fluid dynamics. In this talk I'll explain how recent data science results in non-metric multidimensional scaling provide a new perspective on the Hawking-Malament theorem that is the foundation of the causal set program for quantum gravity. I'll describe a new algorithm for embedding causal sets in Lorentzian manifolds motivated by this perspective. And I'll end with some speculations about possible quantum dynamics for causal sets. Familiarity with the causal set program for quantum gravity will not be assumed.