Cyrus Neary
Papers
2
Total Citations
24
H-Index
2
About
Cyrus Neary is an emerging researcher at the intersection of machine learning and dynamical systems, with a focus on physics-informed neural networks and data-efficient learning for complex systems. His work addresses one of the central challenges in applied machine learning: how to build models that are both accurate and generalizable when training data is scarce. Neary's most cited contribution, "Neural Networks with Physics-Informed Architectures and Constraints for Dynamical Systems Modeling" (2021, 22 citations), demonstrates how embedding prior physical knowledge — such as conservation laws and system Jacobians — directly into neural network architectures can dramatically improve data efficiency and generalization in dynamical systems modeling. Building on this foundation, his 2023 work on physics-constrained neural stochastic differential equations pushes the boundary further, presenting frameworks capable of learning reliable controlled dynamics models from as little as three minutes of observed data, while also quantifying uncertainty — a critical feature for safe real-world deployment. Neary's research is particularly relevant to robotics, control systems, and scientific computing communities, where data collection is expensive and model reliability is paramount. His blend of physical first principles with modern deep learning techniques positions him as a thoughtful contributor to the growing field of scientific machine learning.
Research Focus
Key Achievements
Top Papers
- 1
- 2