An Investigation of Machine Learning Methods Applied to Structure Prediction in Condensed Matter
Authors:
- William J. BrouwerWBWilliam J. Brouwer
- C. Lee GilesCGC. Lee Giles
Preprint
Abstract
Materials characterization remains a significant, time-consuming undertaking. Generally speaking, spectroscopic techniques are used in conjunction with empirical and ab-initio calculations in order to elucidate structure. These experimental and computational methods typically require significant human input and interpretation, particularly with regards to novel materials. Recently, the application of data mining and machine learning to problems in material science have shown great promise in reducing this overhead. In the work presented here, several aspects of machine learning are explored with regards to characterizing a model material, titania, using solid-state Nuclear Magnetic Resonance (NMR). Specifically, a large dataset is generated, corresponding to NMR $^{47}$Ti spectra, using ab-initio calculations for generated TiO$_2$ structures. Principal Components Analysis (PCA) reveals that input spectra may be compressed by more than 90%, before being used for subsequent machine learning. Two key methods are used to learn the complex mapping between structural details and input NMR spectra, demonstrating excellent accuracy when presented with test sample spectra. This work compares Support Vector Regression (SVR) and Artificial Neural Networks (ANNs), as one step towards the construction of an expert system for solid state materials characterization.
Keywords
All Related Versions
This publication has 23 references indexed in Scilit:
- AFLOWLIB.ORG: A distributed materials properties repository from high-throughput ab initio calculationsComputational Materials Science, 2012
- A search model for topological insulators with high-throughput robustness descriptorsNature Materials, 2012
- Crystal structure prediction for iron as inner core material in heavy terrestrial planetsEarth and Planetary Science Letters, 2011
- Adsorption of Zn2+ on the (110) Surface of TiO2 (Rutile): A Density Functional Molecular Dynamics StudyThe Journal of Physical Chemistry C, 2011
- Optimized multiple quantum MAS lineshape simulations in solid state NMRComputer Physics Communications, 2009
- Structure Elucidation from 2D NMR Spectra Using the StrucEluc Expert System: Detection and Removal of Contradictions in the DataJournal of Chemical Information and Computer Sciences, 2004
- Local structure in perovskite relaxor ferroelectrics: high-resolution 93Nb 3QMAS NMRJournal of Magnetic Resonance, 2004
- A Tutorial on Support Vector Machines for Pattern RecognitionData Mining and Knowledge Discovery, 1998
- Automated analysis of protein NMR assignments using methods from artificial intelligenceJournal of Molecular Biology, 1997
- Stock performance modeling using neural networks: A comparative study with regression modelsNeural Networks, 1994