18,586 Hits in 5.2 sec

Measuring the Similarity between Materials with an Emphasis on the Materials Distinctiveness [article]

Tran-Thai Dang, Tien-Lam Pham, Hiori Kino, Takashi Miyake, and Hieu-Chi Dam
2019 arXiv   pre-print
Empirical experiments and theoretical analysis reveal that similarity measures and kernels that minimize the loss of materials distinctiveness improve the prediction performance.  ...  In this study, we establish a basis for selecting similarity measures when applying machine learning techniques to solve materials science problems.  ...  Acknowledgements This work was partly supported by PRESTO and the "Materials Research by Information Integration" initiative (MI 2 I) project of the Support Program for Starting Up Innovation Hub, by the  ... 
arXiv:1903.10867v1 fatcat:ilm5uskelvan3d6rv4fboqag4q

Band gap prediction for large organic crystal structures with machine learning [article]

Bart Olsthoorn, R. Matthias Geilhufe, Stanislav S. Borysov, Alexander V. Balatsky
2018 arXiv   pre-print
Finally, the trained models are employed to predict the band gap for 260,092 materials contained within the Crystallography Open Database (COD) and made available online so the predictions can be obtained  ...  For example, by visualizing the SOAP kernel similarity between the crystals, different clusters of materials can be identified, such as organic metals or semiconductors.  ...  Acknowledgement We are grateful for support from the Swedish  ... 
arXiv:1810.12814v3 fatcat:ahe54ivmlnchbf7nndelkqcl5i

How to represent crystal structures for machine learning: Towards fast prediction of electronic properties

K. T. Schütt, H. Glawe, F. Brockherde, A. Sanna, K. R. Müller, E. K. U. Gross
2014 Physical Review B  
As an alternative, we here propose a machine learning approach for the fast prediction of solid-state properties. To achieve this, LSDA calculations are used as training set.  ...  We propose a novel crystal structure representation for which learning and competitive prediction accuracies become possible within an unrestricted class of spd systems.  ...  The property to be predicted is computed as a linear combination of kernel functions of the material of interest and the training materials.  ... 
doi:10.1103/physrevb.89.205118 fatcat:msemwzpfwregxmrvd3ng6s2b4q

Unified Representation of Molecules and Crystals for Machine Learning [article]

Haoyan Huo, Matthias Rupp
2018 arXiv   pre-print
Empirical evidence is presented for energy prediction errors below 1 kcal/mol for 7k organic molecules and 5 meV/atom for 11k elpasolite crystals.  ...  For this, kernel learning approaches crucially require a single Hilbert space accommodating arbitrary atomistic systems.  ...  Kernel-based ML models [9] for fast accurate prediction of ab initio properties require a single Hilbert space of atomistic systems in which regression is carried out.  ... 
arXiv:1704.06439v3 fatcat:mon7sqsekbhjriuqrh5ezcsoty

Comparing molecules and solids across structural and alchemical space

Sandip De, Albert P. Bartók, Gábor Csányi, Michele Ceriotti
2016 Physical Chemistry, Chemical Physics - PCCP  
A general procedure to compare molecules and materials powers insightful representations of energy landscapes and precise machine-learning predictions of properties.  ...  Several other descriptors have been also used in machine learning, to predict properties of materials and molecules circumventing the need for an expensive electronic structure calculation.  ...  of physical-chemical properties of materials and molecules.  ... 
doi:10.1039/c6cp00415f pmid:27101873 fatcat:rs543j6ukfdnjk7ysulcqktat4

Orbital Graph Convolutional Neural Network for Material Property Prediction [article]

Mohammadreza Karamad, Rishikesh Magar, Yuting Shi, Samira Siahrostami, Ian D. Gates, Amir Barati Farimani
2020 arXiv   pre-print
Material representations that are compatible with machine learning models play a key role in developing models that exhibit high accuracy for property prediction.  ...  Therefore, to develop robust machine learningmodels for material properties prediction, it is imperative to include features representing such chemical attributes.  ...  Acknowledgements The template for the preprint has been taken from: https://  ... 
arXiv:2008.06415v1 fatcat:bzjq54pmwfcidgt5r4y7ddutiq

Machine Learning, Quantum Mechanics, and Chemical Compound Space [article]

Raghunathan Ramakrishnan, O. Anatole von Lilienfeld
2016 arXiv   pre-print
The models are trained and validated using standard quantum chemistry results obtained for organic molecules and materials selected from chemical space at random.  ...  We review recent studies dealing with the generation of machine learning models of molecular and solid properties.  ...  All aforementioned local atomic properties were modeled using Laplacian kernel and Manhattan distance.  ... 
arXiv:1510.07512v3 fatcat:5wulan4e6zck3k3jfuxci2fl4a

Investigating 3D Atomic Environments for Enhanced QSAR [article]

William McCorkindale, Carl Poelking, Alpha A. Lee
2020 arXiv   pre-print
Predicting bioactivity and physical properties of molecules is a longstanding challenge in drug design.  ...  We describe a novel alignment-free 3D QSAR method using Smooth Overlap of Atomic Positions (SOAP), a well-established formalism developed for interpolating potential energy surfaces.  ...  Acknowledgements We thank Gábor Csányi for insightful discussion. WM acknowledges the support of the Gates Cambridge Trust. AAL acknowledges the Winton Programme for the Physics of Sustainability.  ... 
arXiv:2010.12857v1 fatcat:noe3h2mf6bfurf5g7rfp6sgope

Representations of molecules and materials for interpolation of quantum-mechanical simulations via machine learning [article]

Marcel F. Langer, Alex Goeßmann, Matthias Rupp
2021 arXiv   pre-print
For selected state-of-the-art representations, we compare energy predictions for organic molecules, binary alloys, and Al-Ga-In sesquioxides in numerical experiments controlled for data distribution, regression  ...  Computational study of molecules and materials from first principles is a cornerstone of physics, chemistry, and materials science, but limited by the cost of accurate and precise simulations.  ...  The horizontal axis represents chemical or materials space, the vertical axis the predicted property.  ... 
arXiv:2003.12081v2 fatcat:wklvbipn65au7mwqbypkynfxw4

Learning with Graph Kernels in the Chemical Universe

Yu-Hang Tang
2019 Zenodo  
Presentations slides of Yu-Hang Tang on application of active machine learning and graph kernels. The talk also features the release of the GraphDot library.  ...  variance Summary › Active learning using GPR can be powerful for predicting molecular properties xyz')) kernel([g1], nodal=True) Examples: 3D molecules node base kernel: element edge base kernel  ...  Phys. 146, 114107 (2017) Gaussian Process Regression using the Marginalized Graph Kernel 9/17/2019 Yu-› For example, a Gaussian adjacency rule = exp − 1 2 − 2 2 › length/distance,  ... 
doi:10.5281/zenodo.3364077 fatcat:s6plpwwb7bb6batcnau56idecy

Data Science Based Mg Corrosion Engineering

Tim Würger, Christian Feiler, Félix Musil, Gregor B. V. Feldbauer, Daniel Höche, Sviatlana V. Lamaka, Mikhail L. Zheludkevich, Robert H. Meißner
2019 Frontiers in Materials  
As the sheer number of potential dissolution modulators makes it impossible to obtain a detailed atomistic understanding of the inhibition mechanisms for each additive, other measures for inhibition prediction  ...  interface properties.  ...  Ceriotti from École Polytechnique Fédérale de Lausanne, Switzerland are acknowledged for discussions about setting up machine learning for magnesium dissolution modulators.  ... 
doi:10.3389/fmats.2019.00053 fatcat:bhzxaduxpbgipklvyxktrehb4q

Building nonparametric n-body force fields using Gaussian process regression [article]

Aldo Glielmo, Claudio Zeni, Ádám Fekete, Alessandro De Vita
2019 arXiv   pre-print
The models automatically selected for the two materials were found to be in agreement with physical intuition.  ...  For accurate regression it is fundamental to incorporate prior knowledge into the GP kernel function.  ...  We, AG, CZ, and AF, are immensely grateful to Alessandro De Vita for having devoted, with inexhaustible energy and passion, an extra-ordinary amount of his time and brilliance towards our personal and  ... 
arXiv:1905.07626v1 fatcat:ordxu6qztrbifpowcodwbjd6m4

Wasserstein metric for improved QML with adjacency matrix representations [article]

Onur Çaylak, O. Anatole von Lilienfeld, Björn Baumeier
2020 arXiv   pre-print
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.  ...  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  ...  By now, QML has become a viable and popular tool for generating surrogate property models enabling rapid estimates of relevant molecular and materials properties, holding great promise for computational  ... 
arXiv:2001.11005v1 fatcat:4oncn32mfbcz3lghvfydvhuzkq

Machine Learning of Atomic-Scale Properties Based on Physical Principles [chapter]

Michele Ceriotti, Michael J. Willatt, Gábor Csányi
2018 Handbook of Materials Modeling  
We also discuss recent generalisations that allow fine control of correlations between different atomic species, prediction and fitting of tensorial properties, and also how to construct structural kernels  ...  We briefly summarize the kernel regression approach, as used recently in materials modelling, to fitting functions, particularly potential energy surfaces, and highlight how the linear algebra framework  ...  Linear combination of kernels When fitting interatomic potentials for materials, a model is constructed for the atomic energy, sometimes called the "site energy".  ... 
doi:10.1007/978-3-319-42913-7_68-1 fatcat:cd3ikye5a5fkpps73yief6csnq

Reactivity of amorphous carbon surfaces: rationalizing the role of structural motifs in functionalization using machine learning

Miguel A. Caro, Anja Aarva, Volker L. Deringer, Gábor Csányi, Tomi Laurila
2018 Chemistry of Materials  
Systematic atomistic studies of surface reactivity for amorphous materials have not been possible in the past because of the complexity of these materials and the lack of the computer power necessary to  ...  This methodology allows us to process large amounts of atomic data to classify carbon atomic motifs on the basis of their geometry and quantify their reactivity toward hydrogen-and oxygen-containing functionalities  ...  ML Model and Kernel Optimization. Our ML model for adsorption energy prediction is a GAP model, described in detail in refs 38 and 39.  ... 
doi:10.1021/acs.chemmater.8b03353 pmid:30487663 pmcid:PMC6251556 fatcat:4f5o46w63zh3bch6ykq3m73zzy
« Previous Showing results 1 — 15 out of 18,586 results