Gaussian Process Regression for Geometry Optimization

We implemented a geometry optimizer based on Gaussian process regression (GPR) to find minimum structures on potential energy surfaces. We tested both a two times differentiable form of the Matern kernel and the squared exponential kernel. The Matern kernel performs much better. We give a detailed description of the optimization procedures. These include overshooting the step resulting from GPR in order to obtain a higher degree of interpolation vs. extrapolation. In a benchmark against the L-BFGS optimizer of the DL-FIND library on 26 test systems, we found the new optimizer to generally reduce the number of required optimization steps.

[1]  S. C. Rogers,et al.  QUASI: A general purpose implementation of the QM/MM approach and its application to problems in catalysis , 2003 .

[2]  J. Perdew,et al.  Density-functional approximation for the correlation energy of the inhomogeneous electron gas. , 1986, Physical review. B, Condensed matter.

[3]  C. G. Broyden.  The Convergence of a Class of Double-rank Minimization Algorithms 1. General Considerations , 1970 .

[4]  Eamonn F. Healy,et al.  Development and use of quantum mechanical molecular models. 76. AM1: a new general purpose quantum mechanical molecular model , 1985 .

[5]  Douglas B Kell,et al.  Optimal construction of a fast and accurate polarisable water potential based on multipole moments trained by machine learning. , 2009, Physical chemistry chemical physics : PCCP.

[6]  Alexey A. Sokol,et al.  ChemShell—a modular software package for QM/MM simulations , 2014 .

[7]  M. Frisch,et al.  Efficient Geometry Minimization and Transition Structure Optimization Using Interpolated Potential Energy Surfaces and Iteratively Updated Hessians. , 2017, Journal of chemical theory and computation.

[8]  Jon Baker,et al.  The location of transition states: A comparison of Cartesian, Z‐matrix, and natural internal coordinates , 1996 .

[9]  F. Weigend,et al.  Balanced basis sets of split valence, triple zeta valence and quadruple zeta valence quality for H to Rn: Design and assessment of accuracy. , 2005, Physical chemistry chemical physics : PCCP.

[10]  P. Popelier,et al.  The prediction of atomic kinetic energies from coordinates of surrounding atoms using kriging machine learning , 2014, Theoretical Chemistry Accounts.

[11]  J. Nocedal.  Updating Quasi-Newton Matrices With Limited Storage , 1980 .

[12]  Walter Thiel,et al.  Structure-based sampling and self-correcting machine learning for accurate calculations of potential energy surfaces and vibrational levels. , 2017, The Journal of chemical physics.

[13]  O. Anatole von Lilienfeld,et al.  Machine Learning, Quantum Chemistry, and Chemical Space , 2017 .

[14]  Walter Thiel,et al.  DL-FIND: an open-source geometry optimizer for atomistic simulations. , 2009, The journal of physical chemistry. A.

[15]  Zhenwei Li,et al.  Molecular dynamics with on-the-fly machine learning of quantum-mechanical forces. , 2015, Physical review letters.

[16]  Jorge Nocedal,et al.  On the limited memory BFGS method for large scale optimization , 1989, Math. Program..

[17]  M. Hestenes,et al.  Methods of conjugate gradients for solving linear systems , 1952 .

[18]  R. Fletcher,et al.  A New Approach to Variable Metric Algorithms , 1970, Comput. J..

[19]  Carl E. Rasmussen,et al.  Gaussian processes for machine learning , 2005, Adaptive computation and machine learning.

[20]  A. Becke,et al.  Density-functional exchange-energy approximation with correct asymptotic behavior. , 1988, Physical review. A, General physics.

[21]  Raghunathan Ramakrishnan,et al.  Many Molecular Properties from One Kernel in Chemical Space. , 2015, Chimia.

[22]  Mills,et al.  Quantum and thermal effects in H2 dissociative adsorption: Evaluation of free energy barriers in multidimensional quantum systems. , 1994, Physical review letters.

[23]  K. Müller,et al.  Machine Learning Predictions of Molecular Properties: Accurate Many-Body Potentials and Nonlocality in Chemical Space , 2015, The journal of physical chemistry letters.

[24]  G. Henkelman,et al.  A climbing image nudged elastic band method for finding saddle points and minimum energy paths , 2000 .

[25]  David P Tew,et al.  Efficient and accurate evaluation of potential energy matrix elements for quantum dynamics using Gaussian process regression. , 2016, The Journal of chemical physics.

[26]  D. Goldfarb.  A family of variable-metric methods derived by variational means , 1970 .

[27]  R. Kondor,et al.  On representing chemical environments , 2012, 1209.3140.

[28]  P. Popelier,et al.  Intramolecular polarisable multipolar electrostatics from the machine learning method Kriging , 2011 .

[29]  Aki Vehtari,et al.  Nudged elastic band calculations accelerated with Gaussian process regression. , 2017, The Journal of chemical physics.

[30]  J. Behler.  Perspective: Machine learning potentials for atomistic simulations. , 2016, The Journal of chemical physics.

[31]  R. Fletcher.  Practical Methods of Optimization , 1988 .