MLatom

Machine learning for atomistic simulations

A comparative study of diﬀerent machine learning methods for dissipative quantum dynamics

Luis E. Herrera Rodríguez, Arif Ullah, Kennet J. Rueda Espinosa, Pavlo O. Dral*, Alexei A. Kananenka*. A comparative study of different machine learning methods for dissipative quantum dynamics. Mach. Learn. Sci. Technol. 2022, accepted. DOI: 10.1088/2632-2153/ac9a9d.

Book “Quantum Chemistry in the Age of Machine Learning”

AIQM1: Artificial Intelligence–Quantum Mechanical Method 1

Peikun Zheng, Roman Zubatyuk, Wei Wu, Olexandr Isayev, Pavlo O. Dral, Artificial Intelligence-Enhanced Quantum Chemical Method with Broad Applicability, Nat. Commun., 2021, 12, 7022. DOI: 10.1038/s41467-021-27340-2.

AIQM1
Accurate enthalpies of formation
Interfaces to popular machine learning potentials (sGDML, GAP-SOAP, ANI, PhysNet, DPMD)
Converting XYZ coordinates to molecular descriptors (RE descriptor, Coulomb matrix)
Sampling (structure-based, farthest-point, random, user-defined)
Δ-ML
Kernel ridge regression (Gaussian, Laplacian, exponential, Matérn kernels; optimization of hyperparameters)
Creating and re-using ML models
Estimating accuracy of ML models (cross-validation etc.)
First derivatives
ML-nuclear ensemble approach for absorption spectrum calculation
learning curves
hyperparameter optimization