Efficient hyperparameter tuning for kernel ridge regression with Bayesian optimization

Machine learning methods usually depend on internal parameters -- so called hyperparameters -- that need to be optimized for best performance. Such optimization poses a burden on machine learning practitioners, requiring expert knowledge, intuition or computationally demanding brute-force parameter searches. We here address the need for more efficient, automated hyperparameter selection with Bayesian optimization. We apply this technique to the kernel ridge regression machine learning method for two different descriptors for the atomic structure of organic molecules, one of which introduces its own set of hyperparameters to the method. We identify optimal hyperparameter configurations and infer entire prediction error landscapes in hyperparameter space, that serve as visual guides for the hyperparameter dependence. We further demonstrate that for an increasing number of hyperparameters, Bayesian optimization becomes significantly more efficient in computational time than an exhaustive grid search -- the current default standard hyperparameter search method -- while delivering an equivalent or even better accuracy.

[1]  Gideon S. Mann,et al.  Efficient Transfer Learning Method for Automatic Hyperparameter Tuning , 2014, AISTATS.

[2]  Jukka Corander,et al.  Bayesian inference of atomistic structure in functional materials , 2017, npj Computational Materials.

[3]  Robert P. Sheridan,et al.  Deep Neural Nets as a Method for Quantitative Structure-Activity Relationships , 2015, J. Chem. Inf. Model..

[4]  A. Choudhary,et al.  Perspective: Materials informatics and big data: Realization of the “fourth paradigm” of science in materials science , 2016 .

[5]  Geun Ho Gu,et al.  Machine learning for renewable energy materials , 2019, Journal of Materials Chemistry A.

[6]  Pavlo O. Dral,et al.  Quantum chemistry structures and properties of 134 kilo molecules , 2014, Scientific Data.

[7]  Kieron Burke,et al.  Guest Editorial: Special Topic on Data-Enabled Theoretical Chemistry. , 2018, The Journal of chemical physics.

[8]  Ramakrishnan Kannan,et al.  HyperSpace: Distributed Bayesian Hyperparameter Optimization , 2018, 2018 30th International Symposium on Computer Architecture and High Performance Computing (SBAC-PAD).

[9]  G. Hunault,et al.  Dataset’s chemical diversity limits the generalizability of machine learning predictions , 2019, Journal of Cheminformatics.

[10]  Samuel B. Williams,et al.  ASSOCIATION FOR COMPUTING MACHINERY , 2000 .

[11]  Michael U. Gutmann,et al.  Bayesian Optimization for Likelihood-Free Inference of Simulator-Based Statistical Models , 2015, J. Mach. Learn. Res..

[12]  K. Müller,et al.  Fast and accurate modeling of molecular atomization energies with machine learning. , 2011, Physical review letters.

[13]  Klavs F. Jensen,et al.  Autonomous discovery in the chemical sciences part I: Progress , 2020, Angewandte Chemie.

[14]  A. Zunger Inverse design in search of materials with target functionalities , 2018 .

[15]  Matthias Seeger,et al.  Learning search spaces for Bayesian optimization: Another view of hyperparameter transfer learning , 2019, NeurIPS.

[16]  Tony Hey,et al.  The Fourth Paradigm: Data-Intensive Scientific Discovery , 2009 .

[17]  Johannes T. Margraf,et al.  Atomic structures and orbital energies of 61,489 crystal-forming organic molecules , 2020, Scientific Data.

[18]  Connor W. Coley,et al.  Autonomous discovery in the chemical sciences part II: Outlook , 2020, Angewandte Chemie.

[19]  Ryan P. Adams,et al.  Design of efficient molecular organic light-emitting diodes by a high-throughput virtual screening and experimental approach. , 2016, Nature materials.

[20]  M. Marques,et al.  Recent advances and applications of machine learning in solid-state materials science , 2019, npj Computational Materials.

[21]  Christopher J. Bartel,et al.  Machine learning for heterogeneous catalyst design and discovery , 2018 .

[22]  Jay F. Whitacre,et al.  The Materials Research Platform: Defining the Requirements from User Stories , 2019, Matter.

[23]  Markus Schneider,et al.  First-principles data set of 45,892 isolated and cation-coordinated conformers of 20 proteinogenic amino acids , 2015, Scientific Data.

[24]  Matthias Rupp,et al.  Unified representation of molecules and crystals for machine learning , 2017, Mach. Learn. Sci. Technol..

[25]  P. Rinke,et al.  Data‐Driven Materials Science: Status, Challenges, and Perspectives , 2019, Advanced science.

[26]  Matthias Rupp,et al.  Machine learning for quantum mechanics in a nutshell , 2015 .

[27]  Hang Lei,et al.  Hyperparameter Optimization for Machine Learning Models Based on Bayesian Optimization , 2019 .

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

[29]  Gowoon Cheon,et al.  Machine learning-assisted discovery of many new solid Li-ion conducting materials , 2018, 1808.02470.

[30]  Andreas Krause,et al.  Information-Theoretic Regret Bounds for Gaussian Process Optimization in the Bandit Setting , 2009, IEEE Transactions on Information Theory.

[31]  Raynald Gauvin,et al.  Application of machine learning methods for the prediction of crystal system of cathode materials in lithium-ion batteries , 2016 .

[32]  M. Rupp,et al.  Chemical diversity in molecular orbital energy predictions with kernel ridge regression. , 2018, The Journal of chemical physics.

[33]  Haoyan Huo,et al.  Unified Representation for Machine Learning of Molecules and Crystals , 2017 .

[34]  Randal S. Olson,et al.  Evaluation of a Tree-based Pipeline Optimization Tool for Automating Data Science , 2016, GECCO.