ThermoEPred-EL: Robust bandgap predictions of chalcogenides with diamond-like structure via feature cross-based stacked ensemble learning

Wang, Xiangmeng; Xu, Yonglin; Yang, Jiong; Ni, Jianyue; Wu, Zhaocong; Zhu, Wenhao

doi:10.1016/j.commatsci.2019.109117

Cited by 18 publications

(7 citation statements)

References 63 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Chalcogenides, especially the binary ones with ZnS-type structure, are promising candidates. Transition metals (most common components in binary ZnS-type chalcogenides) are prone to statistical mixing, which allows for a precise control of the band structure in these materials [163]. Ternary chalcogenides were predicted to be promising p-type transparent conductors [162].…”

Section: Property Prediction Using Machine Learning For Transition Me...mentioning

confidence: 99%

Green Chemistry Applied to Transition Metal Chalcogenides through Synthesis, Design of Experiments, Life Cycle Assessment, and Machine Learning

Pinto¹,

Cho²,

Oliynyk³

et al. 2022

Green Chemistry - New Perspectives

View full text Add to dashboard Cite

Transition metal chalcogenides (TMC) is a broad class of materials comprising binary, ternary, quaternary, and multinary oxides, sulfides, selenides, and tellurides. These materials have application in different areas such as solar cells, photocatalysis, sensors, photoinduced therapy, and fluorescent labeling. Due to the technological importance of this class of material, it is necessary to find synthetic methods to produce them through procedures aligned with the Green Chemistry. In this sense, this chapter presents opportunities to make the solution chemistry synthesis of TMC greener. In addition to synthesis, the chapter presents different techniques of experimental planning and analysis, such as design of experiments, life cycle assessment, and machine learning. Then, it explains how Green Chemistry can benefit from each one of these techniques, and how they are related to the Green Chemistry Principles. Focus is placed on binary chalcogenides (sulfides, selenides, and tellurides), and the quaternary sulfide Cu2ZnSnS4 (CZTS), due to its application in many fields like solar energy, photocatalysis, and water splitting. The Green Chemistry synthesis, characterization, and application of these materials may represent sustainable and effective ways to save energy and resources without compromising the quality of the produced material.

show abstract

Section: Property Prediction Using Machine Learning For Transition Me...mentioning

confidence: 99%

Green Chemistry Applied to Transition Metal Chalcogenides through Synthesis, Design of Experiments, Life Cycle Assessment, and Machine Learning

Pinto¹,

Cho²,

Oliynyk³

et al. 2022

Green Chemistry - New Perspectives

View full text Add to dashboard Cite

show abstract

“…Models trained on high-fidelity bandgap datasets are more useful. ML algorithms such as least absolute shrinkage and selection operator, support vector regression, gradient boosting decision tree, artificial neural network, deep neural network, and k-nearest neighbor with handcrafted features have been used to establish relationships between composition/structure and high-fidelity bandgap [12][13][14][15][16][17][18]. A chemical formula occasionally corresponds to several structures; therefore, models based on the structural features are more practical.…”

Section: Introductionmentioning

confidence: 99%

Accurate and rapid predictions with explainable graph neural networks for small high-fidelity bandgap datasets

Xiao,

Yang,

Wang

2024

Modelling Simul. Mater. Sci. Eng.

View full text Add to dashboard Cite

Accurate and rapid bandgap prediction is a fundamental task in materials science. We propose graph neural networks with transfer learning to overcome the scarcity of training data for high-fidelity bandgap predictions. We also add a perturbation-based component to our framework to improve explainability. The experimental results show that a framework consisting of graph-level pre-training and standard fine-tuning achieves superior performance on all high-fidelity bandgap prediction tasks and training-set sizes. Furthermore, the framework provides a reliable explanation that considers node features together with the graph structure. We also used the framework to screen 105 potential photovoltaic absorber materials.

show abstract

“…However, both of them are computationally expensive and, thus, are restricted to small systems. As an alternative, the machine learning (ML) method has recently attracted considerable attention for band-gap prediction, which can efficiently manage a huge search space at an extremely low cost [30][31][32][33][34][35][36][37]. For example, by choosing 28 primary atomic properties as input features, Pilania et al [30] obtained a kernel ridge regression (KRR) model to predict the band gaps of 1378 unique double perovskites, where the Pearson correlation coefficient can be as high as 97%.…”

Section: Introductionmentioning

confidence: 99%

“…On top of the 136 compositional properties of inorganic solids, Zhuo et al [31] proposed a support vector regression (SVR) model for gap prediction and found a small root mean square error (RMSE) of 0.45 eV. In addition, by leveraging 42 initial elemental features of chalcogenides, Wang et al [33] developed a stacked ensemble learning (SEL)-gap model with a coefficient of determination (R 2 ) value of 90%. It should be noted that the above-mentioned ML models usually contain a large number of input features, which is actually not beneficial for the high-throughput discovery of desired systems.…”

Section: Introductionmentioning

confidence: 99%

High-Throughput Prediction of the Band Gaps of van der Waals Heterostructures via Machine Learning

Lei

Yuan

et al. 2022

Nanomaterials

View full text Add to dashboard Cite

Van der Waals heterostructures offer an additional degree of freedom to tailor the electronic structure of two-dimensional materials, especially for the band-gap tuning that leads to various applications such as thermoelectric and optoelectronic conversions. In general, the electronic gap of a given system can be accurately predicted by using first-principles calculations, which is, however, restricted to a small unit cell. Here, we adopt a machine-learning algorithm to propose a physically intuitive descriptor by which the band gap of any heterostructures can be readily obtained, using group III, IV, and V elements as examples of the constituent atoms. The strong predictive power of our approach is demonstrated by high Pearson correlation coefficient for both the training (292 entries) and testing data (33 entries). By utilizing such a descriptor, which contains only four fundamental properties of the constituent atoms, we have rapidly predicted the gaps of 7140 possible heterostructures that agree well with first-principles results for randomly selected candidates.

show abstract

ThermoEPred-EL: Robust bandgap predictions of chalcogenides with diamond-like structure via feature cross-based stacked ensemble learning

Cited by 18 publications

References 63 publications

Green Chemistry Applied to Transition Metal Chalcogenides through Synthesis, Design of Experiments, Life Cycle Assessment, and Machine Learning

Green Chemistry Applied to Transition Metal Chalcogenides through Synthesis, Design of Experiments, Life Cycle Assessment, and Machine Learning

Accurate and rapid predictions with explainable graph neural networks for small high-fidelity bandgap datasets

High-Throughput Prediction of the Band Gaps of van der Waals Heterostructures via Machine Learning

Contact Info

Product

Resources

About