Ankit Shrivastava
About me
I’m a Computational and Machine Learning Scientist and currently an applied mathematician at the Computational Science and Mathematics Division of Oak Ridge National Laboratory.
My research centers on advancing and applying techniques in machine learning, signal and image processing, uncertainty quantification, high performance computing and numerical modeling to tackle real-world challenges. I focus on building models that are not only mathematically sound but also impactful in practical settings, often through interdisciplinary collaboration.
My work spans a variety of scientific domains, including materials science, solid mechanics, structural engineering, fluid dynamics, medical imaging and power systems.
I also serve as a reviewer for several leading peer-reviewed journals in the fields of applied mathematics, mechanics, uncertainty quantification, and computational science
Research interests
| Machine Learning | |
| Active Learning (BO / RL) | |
| Physics Informed ML | |
| Multimodal / Multifidelity / Multiscale ML | |
| Manifold Learning | |
| Inverse Problems | |
| Explainable AI (XAI) |
| Computational Modeling | |
| Numerical Methods for PDEs | |
| Multiscale & Multiphysics Modeling | |
| Numerical Linear Algebra | |
| Optimization | |
| High-Performance Computing (HPC) |
Skills and Expertise
Tools
Machine Learning Libraries
-
PyTorch
-
TensorFlow
-
BoTorch
-
R
Software Development
-
Python
-
C
-
C++
High Performance Computing
-
CUDA
-
OpenMP
-
MPI
Simulation & Modeling Tools
-
FEniCS
-
LAMMPS
Models & Algorithms
Neural Networks
- Convolutional Neural Networks (CNNs)
- Autoencoders
- Physics-Informed Neural Networks (PINNs)
- Manifold Learning
Optimization & Learning
- Bayesian Optimization
- Evolutionary Optimization
Data Processing
- PCA (Principal Component Analysis)
- Saliency Mapping Algorithms
