Efficient Inversion of Large-scale Problems Exploiting Structure and Randomization

Dr. Jarom David Hogue
Dr. Jarom David Hogue

Abstract: Randomization is a useful tool for dimensionality reduction and is combined with preconditioning techniques to significantly reduce computational resources required to solve linear systems. This technique is also shown to be highly efficient in combination with dimensionality reduction via Fourier transforms based on Toeplitz structure within the system matrix. In addition, randomization can also be applied to efficiently produce a Kronecker product decomposition.

Speaker’s Bio: Jarom David Hogue (Ph.D. 2020) was born in Orem, Utah in 1983. He received a B.S. degree in mathematics from Arizona State University in 2014, a Master’s degree in applied mathematics from Arizona State University in 2016, and a Ph.D. in applied mathematics from Arizona State University in 2020.  He has authored papers in the area of computational mathematics, and currently has interests in modeling and efficient algorithms for the solution of inverse problems and dimensionality reduction techniques.

Last Updated: November 24, 2020 - 2:09 pm