Computational Mathematics & Statistics Seminar by Aleksei Sorokin: (Quasi-) Monte Carlo Importance Sampling with QMCPy

Speaker:

Aleksei Sorokin, Ph.D., Illinois Institute of Technology

Title:

(Quasi-) Monte Carlo Importance Sampling with QMCPy

Abstract:

Aleksei is in his final semester of the co-terminal program in applied math and data science. He has worked on QMCPy for over 2 years and is planning to continue research into Quasi-Monte Carlo methods in a future graduate program. In this talk, he will be talking about his research. 
(Quasi-)Monte Carlo, (Q)MC, methods are a class of powerful numerical integration methods that have been proven to scale well to high dimensions. Numerous  techniques exist to decrease the computational cost of (Q)MC methods. In this talk he focus on importance sampling, a variation reduction technique that transforms the original problem to be easier for (Q)MC approximation. The build up to composed importance sampling is paralleled by code from his QMCPy package that implements the described concepts.

Computational Mathematics & Statistics