Learn2Hop: Learned Optimization on Rough Landscapes
This work addresses optimization challenges in domains like material design, though it appears incremental as it adapts existing meta-learning techniques to specific problems.
The paper tackled the problem of optimizing non-convex loss surfaces with many local minima, such as in atomic structural optimization, by learning optimization algorithms via meta-learning, resulting in improved efficiency in discovering low-energy minima with promising generalization to new tasks.
Optimization of non-convex loss surfaces containing many local minima remains a critical problem in a variety of domains, including operations research, informatics, and material design. Yet, current techniques either require extremely high iteration counts or a large number of random restarts for good performance. In this work, we propose adapting recent developments in meta-learning to these many-minima problems by learning the optimization algorithm for various loss landscapes. We focus on problems from atomic structural optimization--finding low energy configurations of many-atom systems--including widely studied models such as bimetallic clusters and disordered silicon. We find that our optimizer learns a 'hopping' behavior which enables efficient exploration and improves the rate of low energy minima discovery. Finally, our learned optimizers show promising generalization with efficiency gains on never before seen tasks (e.g. new elements or compositions). Code will be made available shortly.