SyNOF, symbolic regression for reduced density matrix functionals

This project will develop a computational framework that uses machine learning to automatically discover mathematical formulas for describing the electron arrangements of molecules. The electronic structure problem is of great interest in quantum chemistry, where cost-effective methods that capture electron correlation are essential for industrial and technological applications. Current machine learning methods often act as black boxes, making it difficult to understand why they produce certain results. In contrast, this project discovers interpretable equations that researchers can read, understand, and verify. The framework also offers a natural connection to quantum computing: near-term quantum devices can provide electron distribution data that would be used to discover formulas capturing correlation effects beyond the reach of classical computers. The collaboration brings together McMaster University’s expertise in machine learning with CSIC’s knowledge of electronic structure theory, producing open-source software tools, benchmark datasets, and trained researchers with skills spanning computational chemistry, artificial intelligence, and quantum simulation.

Faculty Supervisor:

Rodrigo A. Vargas-Hernández

Student:

Partner:

Institute for Advanced Chemistry of Catalonia (IQAC-CSIC)

Discipline:

Physics

Sector:

Quantum Science

University:

McMaster University

Program: