Special Seminar

---

Contributions to Monte Carlo Methods in Data Science

Speaker: Ajay Jasra, CUHK Shenzhen

Location: Online

Videoconference link: https://nyu.zoom.us/j/97362544155

Date: Monday, November 17, 2025, 7 p.m.

Synopsis:

 In this talk I provide an overview of several projects that have focused upon using Monte Carlo methods in Data Science. Monte Carlo methods are used for the numerical approximation of various formulae found in applied mathematics, statistics and physics. I  focus on two problems (i) parameter estimation associated with partially observed diffusion processes and (ii) Eigenvalue estimation derived from Feynman-Kac formulae in physics. In problem (i) I highlight several methods to provably remove time-discretization bias for parameter estimation. In (ii) I discuss a novel mathematical analysis of existing particle approximation algorithms of lagged Feynman-Kac formulae and how it provides a justification of practitioners' applications of it.

Zoom link: https://nyu.zoom.us/j/97362544155