We propose an efficient meta-algorithm for Bayesian inference problems based on low-degree polynomials, semidefinite programming, and tensor decomposition. The algorithm is inspired by recent lower bound constructions for sum-of-squares and related to the method of moments.

# PACM/Applied Mathematics Colloquium

The PACM Colloquium hosts seminars in the field of applied mathematics. Domains of interest include computational fluid dynamics and material science, dynamical systems, numerical analysis and fast algorithms, stochastic problems and stochastic analysis, graph theory and applications, mathematical biology, financial mathematics and mathematical approaches to signal analysis, image processing and information theory.

For more information about this seminar, contact Amit Singer

**Please click on colloquium title for complete abstract.**

##### Sample-optimal inference, computational thresholds, and the methods of moments

Cornell University

##### Physics in the complex plane

The average quantum physicist on the street would say that a quantum-mechanical Hamiltonian must be Dirac Hermitian (invariant under combined matrix transposition and complex conjugation) in order to guarantee that the energy eigenvalues are real and that time evolution is unitary.

Washington University in St. Louis