NEITALG

Breaking nonconvexity in the optimisation of new learnable iterative algorithms

One of the key challenges in computational mathematics is solving nonconvex optimisation problems. Although traditional iterative algorithms are central to scientific computing, they are typically confined to finding local optima. The ERC-funded NEITALG project aims to develop new, efficient algorithms that can reliably find global solutions to nonconvex functions, backed by rigorous mathematical guarantees. These could pave the way for scientific breakthroughs in areas such as the development of new drugs through the optimisation of molecular properties and the improvement of materials for solar energy. Another goal of NEITALG is to design interpretable and adaptive algorithms with strong generalisation guarantees, contributing to the development of safer and more trustworthy AI.

This project has received funding from the European Research Council (ERC) under the European Union’s Horizon Europe research and innovation programme (Grant Agreement No. 101198055).