Nisha Chandramoorthy

Department of Statistics
Committee on Computational and Applied Mathematics (CCAM)

Nisha Chandramoorthy
Office:
Jones 122B

Background

Previously, she was a James C. Edenfield Early-Career Assistant Professor at the School of Computational Science and Engineering at Georgia Tech. Before that, she was a postdoc at the MIT Institute for Data, Systems and Society. She received her PhD and Master's degrees from MIT and a B. Tech. from IIT Roorkee.

Research

Our research encompasses applied dynamical systems, ergodic theory, scientific computing, and computational statistics. In the past, we developed rigorous yet practical computations related to linear response, which is the smooth change in an invariant measure due to smooth perturbations of a dynamical system. Presently, one focus of our research is on sampling algorithms for Bayesian posteriors, with a particular aim to accelerate inference within dynamical systems.

We have a strong interest in leveraging the dynamical systems approach for optimization, sampling, and learning algorithms for dynamics and operators. We are seeking results that enhance our understanding of how existing algorithms generalize and pave the way for creating more effective ones. Our analyses and algorithm development target computational engineering tasks such as sensitivity analysis, model selection, Bayesian inference and data assimilation, and model order reduction using dynamical data and PDE models, especially those relevant for the geosciences.