We use physics, computation, and interpretable machine learning to study photonic and quantum systems—and to design reliable methods for the next generation of scientific problems.
Led by Mohammad H. Javani, Ph.D.Assistant Professor of Cyber-Physical Systems
Based in Atlanta, GeorgiaClark Atlanta University
Across each field, the goal is the same: find compact, interpretable models and test them against meaningful physical or security benchmarks.
Inverse design, model compression, and interpretable learning for multilayers, metasurfaces, and gratings.
Explore this direction 02 / QUANTUMResource-aware quantum algorithms and careful validation for molecular and physical models.
Explore this direction 03 / DISCOVERYMethods that reveal transferable structure, lower-dimensional representations, and useful design principles.
Explore this direction 04 / TRUSTAuditable quantum cryptography and adaptive defense ideas for cyber-physical systems.
Explore this directionPublished work on pruning shows how to reduce the computational burden in nanophotonic optimization while retaining useful design performance.
View selected publicationsA compact example of the lab’s larger question: how much complexity does a useful scientific model actually need?
Read the paper (opens in new tab)The most useful result is one that survives a stronger baseline, a change in assumptions, and an independent check.
Start with the physics and a clear computational task.
Build simulation data and exact or classical references where possible.
Search for efficient representations and algorithms.
Check accuracy, resources, robustness, and reproducibility.
We welcome thoughtful conversations with students and collaborators working across physics, photonics, computing, and security.