Pouria Behnoudfar

Research Outputs

RL-DAUNCE: Reinforcement Learning-Driven Data Assimilation with Uncertainty-Aware Constrained Ensembles

Bridging Model

Earth System Emulator using an Explanable AI

Supervision and mentorship

Madison Experimental Mathematics Lab (2026):

Project title: Learning Statistically Accurate Dynamics of Complex Systems (Link )

In this project, I supervised four undergraduate students developing a learning algorithm for assimilating high-order moments.

Directed Study (2026)
Madison Experimental Mathematics Lab (2025):

Project title: Integrating ML with DA for complex dynamical systems (Link )

In this project, I supervised four undergraduate students. We developed a framework using ML and DA that addresses critical challenges such as capturing intermittent behaviors and extreme events, often inadequately represented in traditional DA frameworks or physical modeling approaches.

Mentoring a PhD student at the Department of Mathematics, UW-Madison

Collaborations

In addition to my formal appointments, I have held visiting researcher positions at several internationally recognized institutions. These include the School of Computing at the Australian National University; the Oden Institute for Computational Engineering and Sciences at the University of Texas at Austin; the Department of Mathematics at the University of Nottingham; the Department of Mechanical Engineering at the Technical University of Eindhoven; and the Department of Mathematics at Monash University. These experiences have broadened my collaborative network and deepened my engagement with interdisciplinary research across applied mathematics, engineering, and computational sciences.

Research Interest

Constrained operator learning for Data Assimilation.
Multi-model Data Assimilation
Integrating physical models with machine learning techniques to enable digital twins and interpretable machine learning.
High-order time integration for nonlinear dynamical problems.

Page updated

Google Sites

Report abuse