Pranav Bahl

Pranav Bahl

MS Student - University of Michigan, Ann Arbor

Biography

Pranav Bahl is currently a master’s student at the University of Michigan, Ann Arbor where he currently exploring Denoising diffusion probabilistic models (DDPMs) and latent diffusion models utilizing tensor network latent spaces and Data assimilation for dynamical systems via kalman filtering. Previously he completed his master’s degree from the Dept. of Aeronautics, Imperial College London (ICL) in concentration Advanced Computational Methods for Aeronautics, Fluid mechanics and Fluid-structure interaction. At Imperial, he primarily explored autonomous data-driven chaotic dynamical systems utilizing classical and quantum recurrent neural networks. His graduate level courseworks ranged from Turbulence/Turbulence modeling, Multidsciplinary Optimization, Computational fluid dynamics (CFD) to Machine learning/Artificial intelligence, Linear systems theory and Computational linear algebra. He completed his bachelor’s from the Dept. of Mechanical Engineering, Delhi Technological University (DTU). During his bachelor’s degree he was primarily interested in exploring Deep learning based Reduced order models (ROMs) for large scale, physics-based high-fidelity simulations either for autonmous evolution of discretized PDEs or for the state-estimation problems. Pranav’s interests lies at the intersection of Data assimilation, Deep learning and Optimization algorithms, where he aims to enhance interpretability of black-box models with applications pertaining to complex high-order physical phenomena, since scientific computing architectures alone cannot be relied upon to be used with no information of the principles behind the physics of the problems.

       

       

Interests

  • Deep learning and Scientific machine learning (SciML)
  • Data Assimilation via Probabilistic Inference (Ensemble Kalman Filtering)
  • Quantum deep learning and Variational quantum algorithms (VQAs)
  • Numerical methods for Partial differential equations (PDEs)
  • Chaos theory and Dynamical systems
  • Deep learning based Reduced-order modeling (ROM)
  • Turbulence and Computational fluid dynamics (CFD)

Education

MS, Aerospace Engineering, University of Michigan Ann Arbor, 2024-25
Current Research: Generative modelling using Tensor network latent spaces (Recurrent neural networks and Diffusion Models)

MSc, Aerospace Engineering (Fluid Mechanics), Imperial College London (ICL), 2022-23
Concentration: Advanced Computational Methods for Fluid Mechanics
Grade: Distinction

B.Tech, Mechanical Engineering, Delhi Technological University (DTU), 2017-21
Grade: Gpa 8.8 (Out of 10)