Big-Data Analytics in Materials Science Luca M. Ghiringhelli Fritz Haber Institute Hands-on workshop and Humboldt-Kolleg: Density-Functional Theory and Beyond - Basic Principles and Modern Insights Isfahan University of Technology, Isfahan, Iran, May 2 to 13, 2016
64
Embed
BigData Analytics in Materials Science · BigData Analytics in Materials Science Luca M. Ghiringhelli Fritz Haber Institute Handson workshop and HumboldtKolleg: DensityFunctional
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
BigData Analytics in
Materials Science
Luca M. GhiringhelliFritz Haber Institute
Handson workshop and HumboldtKolleg:DensityFunctional Theory and Beyond Basic Principles and Modern Insights
Isfahan University of Technology, Isfahan, Iran, May 2 to 13, 2016
Data, data, data: big data
Bigdata challenge, fourV:Volume (amount of data)Variety (heterogeneity, of form and meaning of data)Veracity (uncertainty of data quality)Velocity ?
Highthroughput screening: query and read out what was stored
Shouldn't we do more?
Analysis
Identify (so far) hidden correlations Identify which materials should be studied next as most promising candidates Identify anomalies
We have a dream
From the periodic table of the elements to a chart of materials
Mendeleev's 1871 periodic table
We have a dream
From the periodic table of the elements to a chart of materials
Mendeleev's 1871 periodic table
Ga=69.7 Ge=72.6
We have a dream
From the periodic table of the elements to a chart of materials:Organize materials according to their properties and functions, e.g.
figure of merit of thermoelectrics (as function of T )
turnover frequency of catalytic materials (as function of T and p)
efficiency of photovoltaic systems
Training setCalculate properties and
functions P, for many materials, iDensityFunctional Theory
Fast PredictionCalculate properties
and functions for new values of d (new materials)
Big Data Analysis
DescriptorFind the appropriate
descriptor di, build a table: | i | di | Pi |
LearningFind the function PSL(d) for the table;
do cross validation.Statistical learning
(Orbital period)² = C (orbit's major axis)³
Learning Discovery→
Suppose to know the trajectories of all planets in the solar system, from accurate observations (experiment)orby numerically integrating general relativity equations (calculations at the highest level of theory)
Data (collected by Tycho Brahe)
Statistical learning(performed by
Johannes Kepler)
Physical law(assessed by
Isaac Newton)
Databases, platforms
“Just Databases”ICSD Inorganic Crystal Structure DB http://icsd.fizkarlsruhe.deCOD Crystallography Open DB http://www.crystallography.net/ESP Electronic Structure Project http://gurka.fysik.uu.se/ESP/CCCBDB Comp. Chemistry Comparison and Benchmark DB
http://cccbdb.nist.gov/Databases + analytic toolsMaterials project http://www.materialsproject.orgAFLOW Atomatic Flow for Materials Discovery http://aflowlib.orgAiiDA Automated interactive infrastructure and DB for Atomistic Simulations
http://www.aiida.netOQMD Open Quantum Materials DB http://oqmd.org/
http://nomadrepository.euhttp://nomadcoe.eu
Codedependent raw data conversion layer codeindependent representation→ →
Outline
Descriptors and fingerprints
A (personaltaste compiled) zoo of machine learning / data mining techniques
Regularised regression
Linear and nonlinear dimensionality reduction
Feature selection
Some words on causal descriptorproperty relationship
Descriptors
Can we predict an optimal material for a complex process (e.g. heterogenous catalysis)
● Valence number Zv● Energy of valence s orbital Es● Energy of valence p orbital Ep● Radius of valence s orbital rs● Radius of valence p orbital rp
Ansatz: atomic features
● Valence number Zv● Energy of valence s orbital Es● Energy of valence p orbital Ep● Radius of valence s orbital rs● Radius of valence p orbital rp
KS
leve
ls [
eV]
Valence p
Valence sRadial probability densities
[Å]
Primary (atomic) features
Radius @ maxAverage radiusTurning point
example: Sn (Tin)
Valence p (HOMO)
Valence s
KS lev els [eV
]
LUMO
(Linear) dimensionality reduction: principal components
Principal component analysis
Pearson, K. "On Lines and Planes of Closest Fit to Systems of Points in Space". Philosophical Magazine 2, 559 (1901)
Orthonormal transformation of coordinates, converting a set of (possibly) linearly correlated coordinates into a new set of linearly uncorrelated (called principal or normal) components, such that the first component has the largest variance and each subsequent has the largest variance constrained to being orthogonal to all the preceding components
Principal component analysis
Pearson, K. "On Lines and Planes of Closest Fit to Systems of Points in Space". Philosophical Magazine 2, 559 (1901)
(Linear) dimensionality reduction: principal components
Focus on “learning”: the algorithm has to improve with data size (“learning by experience”)
What about having a dimensionality reduction, or call it feature selection,
i.e., such that the (best) lowdimensional representation is selected
among (many many) given candidates?
It is time for: compressed sensing
Reference:LMG, J. Vybiral, S. V. Levchenko, C. Draxl, and M. Scheffler,
Phys. Rev. Lett. 114, 105503 (2015)Don't overlook the Supplementary Information!
82 octet AB binary compounds
We have a dreamProof of Concept: Descriptor for the Classification “Zincblende/Wurtzite or Rocksalt?”
Rocksalt
ZincblendeRocksalt/Zincblende
82 octet AB binary compounds
d1
d2 RS
J. A. van Vechten, Phys. Rev. 182, 891 (1969).J. C. Phillips, Rev. Mod. Phys. 42, 317 (1970).J. St. John and A.N. Bloch, Phys. Rev. Lett. 33, 1095 (1974)A. Zunger, Phys. Rev. B 22, 5839 (1980).D. G. Pettifor, Solid State Commun. 51, 31 (1984).Y. Saad, D. Gao, T. Ngo, S. Bobbitt, J. R. Chelikowsky, and W. Andreoni, Phys. Rev. B 85, 104104 (2012).
?
We have a dreamProof of Concept: Descriptor for the Classification “Zincblende/Wurtzite or Rocksalt?”
Rocksalt
ZincblendeRS/ZB
82 octet AB binary compounds
d1
d2 RS
We have a dreamProof of Concept: Descriptor for the Classification “Zincblende/Wurtzite or Rocksalt?”
Ansatz: atomic features
● HOMO● LUMO● Ionization Potential● Electron Affinity● Radius of valence s orbital● Radius of valence p orbital● Radius of valence d orbital● … ?
Ansatz: atomic features
● HOMO● LUMO● Ionization Potential● Electron Affinity● Radius of valence s orbital● Radius of valence p orbital● Radius of valence d orbital● … ?
Regularization (prefer “lower complexity” in the solution)
A more complex regularization:
(Linear) ridge regression
Mathematical formulation of the problem
NP – hard !!!
Mathematical formulation of the problem: sparsity
LASSO: convex problem, equivalent to the NP-hard if features (columns of D) are uncorrelated
LASSO, compressed/ive sensing in Materials Science
Find a descriptor AND an accurate evaluation for the difference in energy between RS and ZB crystal structures for all (82) AB octet semiconductors.ΔE = ΔE ( d )
Possibly identify a 2D descriptor which gives a “nice” representation of the materials in a plane
The task
Ansatz: atomic features
● HOMO● LUMO● Ionization Potential● Electron Affinity● Radius of valence s orbital● Radius of valence p orbital● Radius of valence d orbital● … ?
Ansatz: atomic features
● HOMO● LUMO● Ionization Potential● Electron Affinity● Radius of valence s orbital● Radius of valence p orbital● Radius of valence d orbital● … ?
In practice: formalism borrowed form symbolic regression
Systematic construction of the feature space: EUREQA
Descriptor (candidates: 242)a The largest distance between a H atom and its nearest Si neighborb The shortest distance between a Si atom and its sixthnearest Si neighborc The maximum bond valence sum on a Si atomd The smallest value for the fifthsmallest relative bond length around a Si atome The fourthshortest distance between a Si atom and its eighthnearest neighborf The secondshortest distance between a Si atom and its fifthnearest neighborg The thirdshortest distance between a Si atom and its sixthnearest neighborh The HSi nearestneighbor distance for the hydrogen atom with the fourthsmallest difference between the distances to the two Si atoms nearest to a H atom
T. Müller et al. PRB 89 115202 (2014):Data: ~1000 amorphous structures of 216 Si atoms (saturated)
Property: hole trap depth
EUREQA: genetic programming software. Global optimization (genetic algorithm).Schmidt M., Lipson H., Science, Vol. 324, No. 5923, (2009)
Find a descriptor AND an accurate evaluation for the difference in energy between RS and ZB crystal structures for all (82) AB octet semiconductors.ΔE = ΔE ( d )
Possibly identify a 2D descriptor which gives a “nice” representation of the materials in a plane
The task
Ansatz: atomic features
● HOMO H● LUMO L● Ionization Potential IP● Electron Affinity EA● Radius of valence s orbital rs● Radius of valence p orbital rp● Radius of valence d orbital rd● Thousands of nonlinear
functions of the above
Ansatz: atomic features
● HOMO H● LUMO L● Ionization Potential IP● Electron Affinity EA● Radius of valence s orbital rs● Radius of valence p orbital rp● Radius of valence d orbital rd● Thousands of nonlinear
“Extended” LASSO : features are correlated, so the first 25-30 features selected by lasso when scanning from large to low λ are selected and all single features, all pairs, all triplets... are separately tested via linear regression (the NP-hard problem, but only with 25-30 features)
1D 2D 3D
Finding the descriptor
Twodimensional descriptor
0 0.2 eV 0.45 eV 1.0 eV-0.2 eV
A good model must be predictive within the data domain (interpolation):
cross validation
Performance of the descriptors: accuracy, validation
ε
!
“Complexity”
Erro
r
Training err.
Validation err.
Leave 10% out cross validation
Errors are energies, in eV
Max Absolute Error
Convergence with dimensionality of the descriptor
Regularized regression in practice: do validation
A good model must be predictive within the data domain (interpolation):
cross validation
A better model should be causal:stability analysis
Few words on causality
There are four possibilities (types of causality relationship) behind P(d):
1. d → P : P “listens” to d
2. P → d : d “listens” to P
3. A → d and A → P : There is no direct connection between d and P, but d and P both “listen” to a third “actuator”
4. There is no direct connection between d and P, but they have a common effect (Berkson paradox)
...that listens to both and screams: “I occurred” [Judea Pearl]
[If the admission criteria to a certain graduate school call for either high grades as an undergraduate or special musical talents, then these two attributes will be found to be correlated (negatively) in the student population of that school, even if these attributes are uncorrelated in the population at large (selection bias). Indeed, students with low grades are likely to be exceptionally gifted in music, which explains their admission to graduate school.]
Few words on causality
We are not able to write down a scientific law that connects the descriptor
directly with the total-energy difference between RS and ZB structures.However, ZA, ZB determine these descriptors, and ZA, ZB determine the many-body Hamiltonians and the total-energy difference.
ML has diligently took over “Kepler's work”, but no Newton, yetQuestion: is the latter step always necessary?
Quantitative analysis: effect of noise
The same 2D descriptor is found:
A good model must be predictive within the data domain (interpolation):
cross validation
A better model should be causal:stability analysis
An ideal model should be predictive outside the data domain (extrapolation)
When both carbon diamond and BN are excluded from training:
If all C containing binaries (C, SiC, GeC, and SnC) are excluded from training, i.e. no explicit information on C is given to the model:
Hadn't we known about diamond … we'd have predicted it!
Hadn't we known about any carboncontaining binary …we'd have predicted carbon chemistry (from atomic features)