Abstract
Bayesian Optimization Structure Search (BOSS) is an active machine learning technique for accelerated global exploration of energy and property phase space. It is designed to facilitate machine learning in computational and experimental natural sciences.
BOSS uses Bayesian optimization to build N-dimensional surrogate models for the energy or property landscapes and infer global minima. The models are iteratively refined by sequentially sampling DFT data points that are promising and/or have high information content. In computational structure search, representing heterogenous materials with compact chemical ‘building blocks’ allows us to build in prior knowledge and reduce search dimensionality. The uncertainty-led exploration/exploitation sampling strategy delivers global minima with modest sampling, but also ensures visits to less favorable regions of phase space to gather information on rare events and energy barriers.
BOSS uses Bayesian optimization to build N-dimensional surrogate models for the energy or property landscapes and infer global minima. The models are iteratively refined by sequentially sampling DFT data points that are promising and/or have high information content. In computational structure search, representing heterogenous materials with compact chemical ‘building blocks’ allows us to build in prior knowledge and reduce search dimensionality. The uncertainty-led exploration/exploitation sampling strategy delivers global minima with modest sampling, but also ensures visits to less favorable regions of phase space to gather information on rare events and energy barriers.
| Original language | English |
|---|---|
| Publication status | Published - 9 Mar 2020 |
| MoE publication type | I2 ICT applications |
Keywords
- Software
- Machine learning
- Optimization
- Materials Science
- Computational Physics