arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2106.01138physics.chem-phcs.LGphysics.comp-ph
收藏

基于可微轨迹重新加权的神经网络势函数学习

Learning neural network potentials from experimental data via Differentiable Trajectory Reweighting

Technical University of Munich(慕尼黑工业大学)

Stephan Thaler, Julija Zavadlav

摘要:

在分子动力学(MD)中,基于量子力学数据自底向上训练的神经网络(NN)势最近取得了巨大的成功。直接从实验数据中学习NN势的自顶向下方法受到的关注较少,通常在通过MD模拟反向传播时面临数值和计算挑战。我们提出了微分轨迹加权(DiffTRe)的方法,它绕过微分通过MD模拟的时间无关的可观测量。利用热力学微扰理论,我们避免了梯度爆炸,并在自顶向下学习的梯度计算中实现了大约2个数量级的加速。我们证明了DiffTRe在学习金刚石原子模型和基于不同实验观测值(包括热力学,结构和机械性能)的粗粒度水模型的NN势方面的有效性。重要的是,DiffTRe还将自下而上的结构粗粒度方法(如迭代玻尔兹曼反演)推广到任意电位。所提出的方法构成了一个重要的里程碑,丰富神经网络的潜力与实验数据,特别是当准确的自下而上的数据是不可用的。

英文摘要:

In molecular dynamics (MD), neural network (NN) potentials trained bottom-up on quantum mechanical data have seen tremendous success recently. Top-down approaches that learn NN potentials directly from experimental data have received less attention, typically facing numerical and computational challenges when backpropagating through MD simulations. We present the Differentiable Trajectory Reweighting (DiffTRe) method, which bypasses differentiation through the MD simulation for time-independent observables. Leveraging thermodynamic perturbation theory, we avoid exploding gradients and achieve around 2 orders of magnitude speed-up in gradient computation for top-down learning. We show effectiveness of DiffTRe in learning NN potentials for an atomistic model of diamond and a coarse-grained model of water based on diverse experimental observables including thermodynamic, structural and mechanical properties. Importantly, DiffTRe also generalizes bottom-up structural coarse-graining methods such as iterative Boltzmann inversion to arbitrary potentials. The presented method constitutes an important milestone towards enriching NN potentials with experimental data, particularly when accurate bottom-up data is unavailable.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑