Bayesian Optimization for Materials Science (eBook)

Daniel Packwood (Autor)

eBook Download: PDF

2017 | 1st ed. 2017
VIII, 42 Seiten
Springer Singapore (Verlag)
978-981-10-6781-5 (ISBN)

Infos

This book provides a short and concise introduction to Bayesian optimization specifically for experimental and computational materials scientists. After explaining the basic idea behind Bayesian optimization and some applications to materials science in Chapter 1, the mathematical theory of Bayesian optimization is outlined in Chapter 2. Finally, Chapter 3 discusses an application of Bayesian optimization to a complicated structure optimization problem in computational surface science.

Bayesian optimization is a promising global optimization technique that originates in the field of machine learning and is starting to gain attention in materials science. For the purpose of materials design, Bayesian optimization can be used to predict new materials with novel properties without extensive screening of candidate materials. For the purpose of computational materials science, Bayesian optimization can be incorporated into first-principles calculations to perform efficient, global structure optimizations. While research in these directions has been reported in high-profile journals, until now there has been no textbook aimed specifically at materials scientists who wish to incorporate Bayesian optimization into their own research. This book will be accessible to researchers and students in materials science who have a basic background in calculus and linear algebra.

Inhalt

Preface 6
Contents 8
1 Overview of Bayesian Optimization in Materials Science 10
1.1 Brief Overview of Bayesian Optimisation 10
1.2 Examples of Bayesian Optimisation in Materials Science 13
1.2.1 Prediction of Compounds with Low Thermal Conductivity 13
1.2.2 Prediction of Compounds with Optimal Melting Temperatures and Elastic Properties 14
1.2.3 Prediction of Interface Structures 15
1.2.4 Design of Interface Nanostructure 16
1.3 Bayesian Optimization Requires Good Feature Vectors 17
References 19
2 Theory of Bayesian Optimization 20
2.1 Bayesian Interpretation of Probability 20
2.2 Equilibrium Bond Lengths Via Bayesian Optimization 21
2.2.1 Prior Probability 23
2.2.2 Likelihood Function and Posterior Distribution 24
2.2.3 Example Calculation of the Posterior Distribution 26
2.2.4 The Expected Improvement 27
2.2.5 Example Run of Bayesian Optimisation 28
2.2.6 Training 31
2.3 Bayesian Optimization in the General Case 32
2.4 R Code for Bayesian Optimization 33
Appendix 2.1 34
Appendix 2.2 35
Appendix 2.3 36
References 37
3 Bayesian Optimization of Molecules Adsorbed to Metal Surfaces 38
3.1 Density Functional Theory for Surface Science 38
3.2 Bayesian Optimization for Surface Science 40
3.2.1 Preliminary Computational Study 41
3.2.2 Statement of Optimization Problem 43
3.2.3 Data Description 44
3.2.4 Choice of Feature Vectors 45
3.2.5 Training of Hyperparameters 46
3.2.6 Predictive Performance 48
3.2.7 Discussion 49
References 50

Produkt-Details

• Erscheint lt. Verlag: 4.10.2017
• Reihe/Serie: SpringerBriefs in the Mathematics of Materials
• Reihe/Serie: SpringerBriefs in the Mathematics of Materials
• Zusatzinfo: VIII, 42 p. 16 illus., 12 illus. in color.
• Verlagsort: Singapore
• Sprache: englisch
• Themenwelt: Mathematik / Informatik ► Mathematik ► Angewandte Mathematik
• Themenwelt: Mathematik / Informatik ► Mathematik ► Finanz- / Wirtschaftsmathematik
• Themenwelt: Mathematik / Informatik ► Mathematik ► Statistik
• Themenwelt: Naturwissenschaften ► Physik / Astronomie
• Themenwelt: Technik ► Maschinenbau
• Schlagworte: Bayesian optimisation • catalyst • density functional theory • Intermolecular Interaction • Surface
• ISBN-10: 981-10-6781-3 / 9811067813
• ISBN-13: 978-981-10-6781-5 / 9789811067815