Novick Andrew, Cai Diana, Nguyen Quan, Garnett Roman, Adams Ryan, Toberer Eric
Department of Physics, Colorado School of Mines, Golden, Colorado, USA.
Center for Computational Mathematics, Flatiron Institute Address, New York, New York, USA.
Mater Horiz. 2024 Oct 28;11(21):5381-5393. doi: 10.1039/d4mh00432a.
Active learning is a valuable tool for efficiently exploring complex spaces, finding a variety of uses in materials science. However, the determination of convex hulls for phase diagrams does not neatly fit into traditional active learning approaches due to their global nature. Specifically, the thermodynamic stability of a material is not simply a function of its own energy, but rather requires energetic information from all other competing compositions and phases. Here we present convex hull-aware active learning (CAL), a novel Bayesian algorithm that chooses experiments to minimize the uncertainty in the convex hull. CAL prioritizes compositions that are close to or on the hull, leaving significant uncertainty in other compositions that are quickly determined to be irrelevant to the convex hull. The convex hull can thus be predicted with significantly fewer observations than approaches that focus solely on energy. Intrinsic to this Bayesian approach is uncertainty quantification in both the convex hull and all subsequent predictions (, stability and chemical potential). By providing increased search efficiency and uncertainty quantification, CAL can be readily incorporated into the emerging paradigm of uncertainty-based workflows for thermodynamic prediction.
主动学习是有效探索复杂空间的宝贵工具，在材料科学中有多种用途。然而，由于相图凸包的全局性质，其确定并不完全适用于传统的主动学习方法。具体而言，材料的热力学稳定性不仅仅是其自身能量的函数，而是需要来自所有其他竞争成分和相的能量信息。在此，我们提出了凸包感知主动学习（CAL），这是一种新颖的贝叶斯算法，它选择实验以最小化凸包中的不确定性。CAL优先考虑接近或位于凸包上的成分，而对于那些很快被确定与凸包无关的其他成分则保留显著的不确定性。因此，与仅关注能量的方法相比，通过显著更少的观测就能预测凸包。这种贝叶斯方法的内在特点是在凸包以及所有后续预测（如稳定性和化学势）中进行不确定性量化。通过提高搜索效率和不确定性量化，CAL可以很容易地融入基于不确定性的热力学预测工作流程这一新兴范式中。
