Daria Zrelova
Papers
1
Total Citations
6
H-Index
1
About
Daria Zrelova is a researcher at the forefront of integrating fundamental physics with deep learning, specializing in physically informed neural networks for autonomous systems. Her work centers on developing classical Lagrangian and Hamiltonian neural network models that embed physical conservation laws directly into machine learning architectures. In her most cited paper (2022, 6 citations), Zrelova systematically explores principles for constructing deep learning systems that account for the physical properties of controlled objects, particularly autonomous robots. Her key contribution lies in creating an intelligent toolkit where neural networks are not merely data-driven but are physically aware—incorporating energy conservation, symmetries, and dynamical constraints into their learning processes. This approach significantly improves model generalization, sample efficiency, and physical plausibility compared to traditional black-box neural networks. By bridging the gap between analytical mechanics and modern AI, Zrelova’s work has practical implications for robotics, control theory, and simulation, enabling more robust and interpretable autonomous systems. Her research represents an important step toward machines that understand and respect the laws of physics, making her a notable emerging voice in the physics-informed machine learning community.
Research Focus
Key Achievements
Top Papers
- 1