CoChat

From DFT to machine learning: recent approaches to materials science–a review

Open access

The DOI matches its Crossref record; 508 of the 509 references it cites were found in a registry or an index, and 1 are not checked yet.

Abstract

Recent advances in experimental and computational methods are increasing the quantity and complexity of generated data. This massive amount of raw data needs to be stored and interpreted in order to advance the materials science field. Identifying correlations and patterns from large amounts of complex data is being performed by machine learning algorithms for decades. Recently, the materials science community started to invest in these methodologies to extract knowledge and insights from the accumulated data. This review follows a logical sequence starting from density functional theory as the representative instance of electronic structure methods, to the subsequent high-throughput approach, used to generate large amounts of data. Ultimately, data-driven strategies which include data mining, screening, and machine learning techniques, employ the data generated. We show how these approaches to modern computational materials science are being used to uncover complexities and design novel materials with enhanced properties. Finally, we point to the present research problems, challenges, and potential future perspectives of this new exciting field.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1088/2515-7639/ab084b

Cites

508 of 509 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 Schematic presentation of the topics discussed in this review and their relationships. The possible atomic combinations form a great number of compounds, which can be studied by means of experimental, theoretical or computational approaches, especially with high-throughput calculations. Large quantities of data generated are then stored in databases, which can be used by means of materials screening or machine learning, both of which leads to promising materials candidates. Data-driven or traditional routes select materials suited for speci fi c applications. We illustrate these relationships in the context of two-dimensional materials. Figures from panels adapted with permission from [ 1 ] ( experiments ) ( ©2012MacmillanPublishers Limited. All rights reserved. With permission of Springer ) , [ 2 ] ( computational science and machine learning ) , and [ 3 ] ( materials prediction ) ( ©2018Macmillan Publishers Limited, part of Springer Nature. All rights reserved. With permission of Springer ) .
Figure 1 Schematic presentation of the topics discussed in this review and their relationships. The possible atomic combinations form a great number of compounds, which can be studied by means of experimental, theoretical or computational approaches, especially with high-throughput calculations. Large quantities of data generated are then stored in databases, which can be used by means of materials screening or machine learning, both of which leads to promising materials candidates. Data-driven or traditional routes select materials suited for speci fi c applications. We illustrate these relationships in the context of two-dimensional materials. Figures from panels adapted with permission from [ 1 ] ( experiments ) ( ©2012MacmillanPublishers Limited. All rights reserved. With permission of Springer ) , [ 2 ] ( computational science and machine learning ) , and [ 3 ] ( materials prediction ) ( ©2018Macmillan Publishers Limited, part of Springer Nature. All rights reserved. With permission of Springer ) .CC-BY · From the PDF
Figure 2 Thefourscience paradigms: empirical, theoretical, computational, and data-driven. Each paradigm both bene fi ts from and contributes to the others. Adapted from [ 6 ] . CC BY 4.0
Figure 2 Thefourscience paradigms: empirical, theoretical, computational, and data-driven. Each paradigm both bene fi ts from and contributes to the others. Adapted from [ 6 ] . CC BY 4.0CC-BY · From the PDF
Figure 3 ( Left ) Data Science as an integrative discipline. Adapted from [ 8 ] . ( Right ) Big Data characteristics known as the ' fi ve V ' s ' .
Figure 3 ( Left ) Data Science as an integrative discipline. Adapted from [ 8 ] . ( Right ) Big Data characteristics known as the ' fi ve V ' s ' .CC-BY · From the PDF
Figure 4 Chronological evolution of the number of publications for DFT, HT, ML, and materials informatics. Initial developments of each discipline date to many decades before actual adoption by the community. Data compiled from the Web of Science platform, using each keyword in the ' Topic ' search term.
Figure 4 Chronological evolution of the number of publications for DFT, HT, ML, and materials informatics. Initial developments of each discipline date to many decades before actual adoption by the community. Data compiled from the Web of Science platform, using each keyword in the ' Topic ' search term.CC-BY · From the PDF
Figure 5 Theincreasing gap between data, information, knowledge, and utility, which calls for more ef fi cient approaches to accelerate this conversion. Adapted from [ 21 ] , copyright 2013 with permission from Elsevier.
Figure 5 Theincreasing gap between data, information, knowledge, and utility, which calls for more ef fi cient approaches to accelerate this conversion. Adapted from [ 21 ] , copyright 2013 with permission from Elsevier.CC-BY · From the PDF
Image 6
CC-BY · From the PDF
Image 7
CC-BY · From the PDF
Figure 8 Hierarchical description and techniques examples of arti fi cial intelligence and its machine learning and deep learning subfi elds.
Figure 8 Hierarchical description and techniques examples of arti fi cial intelligence and its machine learning and deep learning subfi elds.CC-BY · From the PDF
Figure 9 Machine learning algorithms and usage diagram, divided into the main types of problems: unsupervised ( dimensionality reduction and clustering ) andsupervised ( classi fi cation and regression ) learning. Adapted from [ 179 ] , copyright © SAS Institute Inc., Cary, NC, USA. All Rights Reserved. Used with permission.
Figure 9 Machine learning algorithms and usage diagram, divided into the main types of problems: unsupervised ( dimensionality reduction and clustering ) andsupervised ( classi fi cation and regression ) learning. Adapted from [ 179 ] , copyright © SAS Institute Inc., Cary, NC, USA. All Rights Reserved. Used with permission.CC-BY · From the PDF
Figure 10 Principal component analysis ( PCA ) performed over a 3D dataset with 3 labels given by the color code ( left ) resulting in a 2Ddataset ( right ) .
Figure 10 Principal component analysis ( PCA ) performed over a 3D dataset with 3 labels given by the color code ( left ) resulting in a 2Ddataset ( right ) .CC-BY · From the PDF
Figure 11 Bias × variance trade-off. The optimum model complexity is evaluated against the prediction error given by the test set. Adapted with permission from [ 185 ] .
Figure 11 Bias × variance trade-off. The optimum model complexity is evaluated against the prediction error given by the test set. Adapted with permission from [ 185 ] .CC-BY · From the PDF
Figure 12 ( a ) Exampleofthe sigmoid function and the classi fi cation of negative ( red ) andpositive ( blue ) examples in logistic regression. The gray arrow points out to the incorrectly classi fi ed points in the dataset. ( b ) AVoronoidiagramdepicting the k -means classi fi cation algorithm. The data point labels correspond to the distinct colors of the scatter points while the assignment to each cluster, de fi ned by their centroids ( black crosses ) , corresponds to the color patches. ( c ) Adendrogramexplaining the hierarchical clustering algorithm, where the color code is a guide to visualize the clusters, and each vertical line represents a cluster. Horizontal lines denote the merging process of two clusters. The number of cuts between a horizontal line and the cluster lines denotes the numberofclusters at a given height, which in the case of the gray dashed line is fi ve. ( d ) k -nearest neighbors Voronoi diagram showing the data point labels and classi fi cation patches as color codes.
Figure 12 ( a ) Exampleofthe sigmoid function and the classi fi cation of negative ( red ) andpositive ( blue ) examples in logistic regression. The gray arrow points out to the incorrectly classi fi ed points in the dataset. ( b ) AVoronoidiagramdepicting the k -means classi fi cation algorithm. The data point labels correspond to the distinct colors of the scatter points while the assignment to each cluster, de fi ned by their centroids ( black crosses ) , corresponds to the color patches. ( c ) Adendrogramexplaining the hierarchical clustering algorithm, where the color code is a guide to visualize the clusters, and each vertical line represents a cluster. Horizontal lines denote the merging process of two clusters. The number of cuts between a horizontal line and the cluster lines denotes the numberofclusters at a given height, which in the case of the gray dashed line is fi ve. ( d ) k -nearest neighbors Voronoi diagram showing the data point labels and classi fi cation patches as color codes.CC-BY · From the PDF
Figure 13 Example of a feed-forward ANN with N hidden layers and a single neuron in the output layer. Red neurons represent sigmoid activated units ( see equation ( 35 )) while yellow ones correspond to the ReLU activation ( equation ( 37 )) .
Figure 13 Example of a feed-forward ANN with N hidden layers and a single neuron in the output layer. Red neurons represent sigmoid activated units ( see equation ( 35 )) while yellow ones correspond to the ReLU activation ( equation ( 37 )) .CC-BY · From the PDF
Image 14
CC-BY · From the PDF
Figure 15 Therevised connectivity between the four science paradigms. From empirical data to fundamental theories, which are realized in computational simulations, generating even more data. Statistical learning in turn can obtain simple phenomenological models that aids in theoretical understanding.
Figure 15 Therevised connectivity between the four science paradigms. From empirical data to fundamental theories, which are realized in computational simulations, generating even more data. Statistical learning in turn can obtain simple phenomenological models that aids in theoretical understanding.CC-BY · From the PDF
Image 16
CC-BY · From the PDF
Image 17
CC-BY · From the PDF
Figure 18 SISSOclassi fi cation of energy differences Δ EAB between rock salt and zinc blende structures of 82 binary AB materials, as arranged by the optimal two-dimensional descriptor. Reproduced from [ 196 ] . CC BY 3.0
Figure 18 SISSOclassi fi cation of energy differences Δ EAB between rock salt and zinc blende structures of 82 binary AB materials, as arranged by the optimal two-dimensional descriptor. Reproduced from [ 196 ] . CC BY 3.0CC-BY · From the PDF
Image 19
CC-BY · From the PDF
Image 20
CC-BY · From the PDF
Image 21
CC-BY · From the PDF

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 509 references

Research works printed but not linked: 7

  1. Eagar, 1995, Technology Review, 98, 42
  2. Bock, 2008, Electron. Journ@l Hist. Probab. Stat., 4, 1
  3. Kohavi, 2002, Decision-tree discovery, Handbook of Data Mining and Knowledge Discovery, 548
  4. Pedregosa, 2011, Journal of Machine Learning Research, 12, 2825
  5. Paszke, 2017, NIPS-W
  6. Li, 2016, Phys. Rev. B, 94, 10.1103/physrevb.94.245129 · DOI
  7. Wang, 2016, Phys. Rev. B, 94, 10.1103/physrevb.94.195105 · DOI · Paper page

Cited by

See all 889

Metadata

Authors
Schleder, G. R., Padilha, A. C. M., Acosta, C. M., Costa, M., & Fazzio, A.
Affiliations
Brazilian Center for Research in Energy and Materials (BR)
Corresponding author
  • Schleder, Gabriel R
  • Fazzio, Adalberto
Venue
Journal of Physics: Materials
Published
2019-05-16
Volume
2
Issue
3
Pages
032001
ISSN
2515-7639 (ISSN-L)
OpenAlex
W2917193084
Cited by
852
References
501
Citations by year
2026: 79 · 2025: 170 · 2024: 135 · 2023: 142 · 2022: 113 · 2021: 112 · 2020: 76 · 2019: 25
FWCI
29.52
Type
journal article
Language
en
Open access
Licence
Locations
Journal of Physics Materials (publishedVersion, cc-by)
Funders
Fundação de Amparo à Pesquisa do Estado de São Paulo (16/14011-2)
Topics
  • Physical Sciences › Materials Science › Materials Chemistry › Machine Learning in Materials Science
  • Physical Sciences › Physics and Astronomy › Atomic and Molecular Physics, and Optics › Advanced Chemical Physics Studies
  • Life Sciences › Pharmacology, Toxicology and Pharmaceutics › Drug Discovery › Computational Drug Discovery Methods
Keywords
machine learning for materials, density functional theory, data-driven materials discovery, data mining, materials screening, electronic structure, materials data

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: From DFT to machine learning: recent approaches to materials science–a review. https://app.cochat.ai/papers/10.1088/2515-7639/ab084b
The paper
Schleder, G. R., Padilha, A. C. M., Acosta, C. M., Costa, M., & Fazzio, A. (2019). From DFT to machine learning: recent approaches to materials science–a review. Journal of Physics: Materials. https://doi.org/10.1088/2515-7639/ab084b

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.