Machine learning of octahedral tilting in oxide perovskites by symbolic classification with compressed sensing

Xie, Stephen; Kotlarz, Parker; Hennig, Richard G.; Nino, Juan C.

doi:10.1016/j.commatsci.2020.109690

Cited by 22 publications

(12 citation statements)

References 57 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Genetic programming has also been used to identify hidden physical laws from the input–output response prior 19,45,46 . We use SISSO because it has been shown to be robust with small amounts of data 40,47,48 . This is advantageous for analysis of CFD systems where the computational cost of simulations increases with increasing number of variables and complexity of the system.…”

Section: Comparison To Related Workmentioning

confidence: 99%

Iterative symbolic regression for learning transport equations

et al. 2022

View full text Add to dashboard Cite

Computational fluid dynamics (CFD) analysis is widely used in chemical engineering. Although CFD calculations are accurate, the computational cost associated with complex systems makes it difficult to obtain empirical equations between system variables. Here, we combine active learning (AL) and symbolic regression (SR) to get a symbolic equation for system variables from CFD simulations. Gaussian process regression‐based AL allows for automated selection of variables by selecting the most instructive points from the available range of possible parameters. The results from these experiments are then passed to SR to find empirical symbolic equations for CFD models. This approach is scalable and applicable for any desired number of CFD design parameters. To demonstrate the effectiveness, we use this method with two model systems. We recover an empirical equation for the pressure drop in a bent pipe and a new equation for predicting backflow in a heart valve under aortic insufficiency.

show abstract

Section: Comparison To Related Workmentioning

confidence: 99%

Iterative symbolic regression for learning transport equations

et al. 2022

View full text Add to dashboard Cite

show abstract

“…Here, we try to find an expression of structural features that best predicts the cation–halogen binding strengths. The particular algorithm used is called SISSO which is based on a compressed sensing technique ( vide infra ). ,− It starts from a set of single-valued elementary features called Φ 0 . It then recursively applies some unary and binary mathematical operations to create new feature spaces called Φ i with i being the number of times the mathematical operations are applied.…”

Section: Theory and Computational Detailsmentioning

confidence: 99%

“…The included features are the effective crystal radii of the metal, halogen, and organic cation, as well as the electronegativity and polarizability of the metal and halogen atoms. The latter two features have shown to be useful in multiple studies combining HOIPs and predictive techniques. , Differently from previous studies and in order to easily encode the nature of the organic cations, here some additional bond counts were included: the number of C–H, N–H and O–H bonds as well as the number of π bonds. Hydroxylammonium is the only cation in the dataset that contains an O–H bond.…”

Section: Theory and Computational Detailsmentioning

confidence: 99%

Molecular Bond Engineering and Feature Learning for the Design of Hybrid Organic–Inorganic Perovskite Solar Cells with Strong Noncovalent Halogen–Cation Interactions

Teunissen

Pieve

2021

J. Phys. Chem. C

View full text Add to dashboard Cite

Hybrid organic–inorganic perovskites are exceedingly interesting candidates for new solar energy technologies for both ground-based and space applications. However, their large-scale production is hampered by the lack of long-term stability, mostly associated with ion migration. The specific role of noncovalent bonds in contributing to the stability remains elusive, and in certain cases controversial. Here, we perform an investigation on a large perovskite chemical space via a combination of first-principles calculations for the bond strengths and the recently developed sure independent screening and sparsifying operator (SISSO) algorithm. The latter is used to formulate mathematical descriptors that, by highlighting the importance of specific noncovalent molecular bonds, can guide the design of perovskites with suppressed ion migration. The results unveil the distinct nature of different noncovalent interactions, with remarkable differences compared to previous arguments and interpretations in the literature on the basis of smaller chemical spaces. In particular, we clarify the origin of the higher stability offered by formamidinium compared to methylammonium, which shows to be different from previous arguments in the literature, and the reasons for the improved stability given by the halogen F and explain the exceptional case of overall stronger bonds for guanidiunium. Within the stability boundaries given by the Goldschmidt factor, the found descriptors give an all-in-one picture of noncovalent interactions which provide more stable configurations, also including interactions other than H bonds. Such descriptors are more informative than previously used quantities and can be used as a universal input to better inform new machine learning studies.

show abstract

“…The latter two features have shown to be useful in multiple studies combining HOIPs and predictive techniques. 88,90 Differently from previous studies and in order to easily encode the nature of the organic cations, here some additional bond counts were included: the number of C-H, N-H and O-H bonds as well as the number of π bonds. Hydroxylammonium is the only cation in the dataset that contains an O-H bond.…”

Section: Sure Independence Screening and Sparsifying Operator (Sisso)...mentioning

confidence: 99%

Molecular Bond Engineering and Feature Learning for the Design of Hybrid Organic-Inorganic Perovskites Solar Cells with Strong Non-Covalent Halogen-Cation Interactions

Teunissen,

da Pieve

2021

Preprint

View full text Add to dashboard Cite

Hybrid organic-inorganic perovskites are exceedingly interesting candidates for new solar energy technologies, for both ground-based and space applications. However, their large-scale production is hampered by the lack of long-term stability, mostly associated to ion migration. The specific role of non-covalent bonds in contributing to the stability remains elusive, and in certain cases controversial. Here, we perform an investigation on a large perovskite chemical space via a combination of first-principles calculations for the bond strengths and the recently developed Sure Independent Screening and Sparsifying Operator (SISSO) algorithm. The latter is used to formulate mathematical descriptors that, by highlighting the importance of specific non-covalent molecular

show abstract

Machine learning of octahedral tilting in oxide perovskites by symbolic classification with compressed sensing

Cited by 22 publications

References 57 publications

Iterative symbolic regression for learning transport equations

Iterative symbolic regression for learning transport equations

Molecular Bond Engineering and Feature Learning for the Design of Hybrid Organic–Inorganic Perovskite Solar Cells with Strong Noncovalent Halogen–Cation Interactions

Molecular Bond Engineering and Feature Learning for the Design of Hybrid Organic-Inorganic Perovskites Solar Cells with Strong Non-Covalent Halogen-Cation Interactions

Contact Info

Product

Resources

About