On-the-fly active learning of interpretable Bayesian force fields for atomistic rare events
On-the-fly active learning of interpretable Bayesian force fields for atomistic rare events
复制标题
DOI：
10.1038/s41524-020-0283-z
复制
发表时间：
2020-03-18
影响因子：
9.7
通讯作者：
Kozinsky, Boris
中科院分区：
文献类型：
--
作者：
Vandermause, Jonathan;Torrisi, Steven B.;Kozinsky, Boris
机器学习力场通常需要手动构建由数千个第一原理计算组成的训练集，当应用于模型训练集中未表示的结构时，这可能会导致训练效率低下和不可预测的错误。这严重限制了这些模型在由重要的罕见事件（例如化学反应和扩散）控制的动力学系统中的实际应用。我们提出了一种自适应贝叶斯推理方法，用于使用分子动力学模拟中动态绘制的结构来自动训练可解释的低维多元素原子间力场。在主动学习框架内，高斯过程回归模型的内部不确定性用于决定是否接受模型预测或执行第一原理计算以增强模型的训练集。该方法适用于一系列单元件和多元件系统，并被证明可以实现精度和计算效率的良好平衡，同时需要最少量的从头开始训练数据。我们提供了我们的方法的完全开源实现，以及将经过训练的模型映射到计算高效的表格力场的程序。
Machine learned force fields typically require manual construction of training sets consisting of thousands of first principles calculations, which can result in low training efficiency and unpredictable errors when applied to structures not represented in the training set of the model. This severely limits the practical application of these models in systems with dynamics governed by important rare events, such as chemical reactions and diffusion. We present an adaptive Bayesian inference method for automating the training of interpretable, low-dimensional, and multi-element interatomic force fields using structures drawn on the fly from molecular dynamics simulations. Within an active learning framework, the internal uncertainty of a Gaussian process regression model is used to decide whether to accept the model prediction or to perform a first principles calculation to augment the training set of the model. The method is applied to a range of single- and multi-element systems and shown to achieve a favorable balance of accuracy and computational efficiency, while requiring a minimal amount of ab initio training data. We provide a fully open-source implementation of our method, as well as a procedure to map trained models to computationally efficient tabulated force fields.