Wasserstein metric for improved QML with adjacency matrix representations

Abstract

We study the Wasserstein metric to measure distances between molecules represented by the atom index dependent adjacency "Coulomb" matrix, used in kernel ridge regression based supervised learning. Resulting quantum machine learning models exhibit improved training efficiency and result in smoother predictions of molecular distortions. We first demonstrate smoothness for the continuous extraction of an atom from some organic molecule. Learning curves, quantifying the decay of the atomization energy's prediction error as a function of training set size, have been obtained for tens of thousands of organic molecules drawn from the QM9 data set. In comparison to conventionally used metrics ($L_1$ and $L_2$ norm), our numerical results indicate systematic improvement in terms of learning curve off-set for random as well as sorted (by norms of row) atom indexing in Coulomb matrices. Our findings suggest that this metric corresponds to a favorable similarity measure which introduces index-invariance in any kernel based model relying on adjacency matrix representations.

References in corpus (11)

• Machine learning for molecular simulation
• Machine Learning Unifies the Modelling of Materials and Molecules
• Towards Exact Molecular Dynamics Simulations with Machine-Learned Force Fields
• Alchemical and structural distribution based representation for improved QML
• Understanding molecular representations in machine learning: The role of uniqueness and target similarity
• sGDML: Constructing Accurate and Data Efficient Molecular Force Fields Using Machine Learning
• Machine learning enables long time scale molecular photodynamics simulations
• Unsupervised machine learning in atomistic simulations, between predictions and understanding
• Guest Editorial: Special Topic on Data-enabled Theoretical Chemistry
• Solid Harmonic Wavelet Scattering for Predictions of Molecule Properties
• Machine learning at the atomic-scale

Cited by in corpus (8)

• Combining Machine Learning and Computational Chemistry for Predictive Insights Into Chemical Systems
• Physics-inspired structural representations for molecules and materials
• Machine learning for electronically excited states of molecules
• Machine learning based energy-free structure predictions of molecules (closed and open-shell), transition states, and solids
• Kernel based quantum machine learning at record rate : Many-body distribution functionals as compact representations
• Are your apps accessible? A GCN-based accessibility checker for low vision users
• Ab initio machine learning in chemical compound space
• A Kernel-based Machine Learning Approach to Computing Quasiparticle Energies within Many-Body Green's Functions Theory