Physical Consistency Bridges Heterogeneous Data in Molecular Multi-Task Learning
This addresses the problem of integrating heterogeneous data in molecular science for researchers, though it is incremental as it builds on existing multi-task learning paradigms with a novel consistency approach.
The paper tackles the challenge of data heterogeneity in molecular multi-task learning by exploiting physical laws to design consistency training approaches, which allow different tasks to exchange information and improve each other, resulting in improved accuracy for structure prediction using more accurate energy data and leveraging force and off-equilibrium structure data.
In recent years, machine learning has demonstrated impressive capability in handling molecular science tasks. To support various molecular properties at scale, machine learning models are trained in the multi-task learning paradigm. Nevertheless, data of different molecular properties are often not aligned: some quantities, e.g. equilibrium structure, demand more cost to compute than others, e.g. energy, so their data are often generated by cheaper computational methods at the cost of lower accuracy, which cannot be directly overcome through multi-task learning. Moreover, it is not straightforward to leverage abundant data of other tasks to benefit a particular task. To handle such data heterogeneity challenges, we exploit the specialty of molecular tasks that there are physical laws connecting them, and design consistency training approaches that allow different tasks to exchange information directly so as to improve one another. Particularly, we demonstrate that the more accurate energy data can improve the accuracy of structure prediction. We also find that consistency training can directly leverage force and off-equilibrium structure data to improve structure prediction, demonstrating a broad capability for integrating heterogeneous data.