Machine learning (ML) has found wide application in materials science. It is believed that a model developed by ML could depict the common trend of the data and therefore reflect the relationship between structure and property, which can be applied to most of the compounds. So, by training ML models with existed databases, important properties of compounds can be predicted ahead of time-consuming experiments or calculations, which will greatly speed up the process of new materials design.
While tremendously useful, these models do not directly show the rules and physics underlying the relationship between structure and property. And despite of their decent overall performance, there will always be some exceptions where ML models fail to give accurate predictions. Very often, it is these exceptions that shed some new insights about the underlying physics, and open up new frontiers in science.
The machine learning results. (a) The scatter plot, and (b) the histogram of errors and the kernel density estimation of the probability density function. Red points and regions correspond to structures with prediction error larger than 2 eV.
Credit: ©Science China Press
A research group led by Prof. Feng Pan, the founding dean of the School of Advanced Materials, Peking University Shenzhen Graduate School, has recently shown that these models are valuable not only when they succeed in predicting properties accurately, but also when they fail.
In their work, a model is built to predict the HSE band gaps of compounds according to their atomic structures only, based on a high-throughput calculation database constructed by the school themselves.
The R2 of the model is 0.89, comparable with similar works. They then filtered out those structures with prediction error larger than 2 eV and examined them carefully.
Many structures with unusual structure units, or showing other abnormities with similar compounds, like relatively large band gaps or being in different phases. Among these unusual structures, AgO2F raises great interest and a detailed analysis is given.
It is found that Ag3+ and O22- coexist in this compound, and while Ag ions are in square planar coordination, there is little hybridization between orbitals of Ag and O. States near the band edges are mainly contributed by O-2p orbitals and the band gap is much smaller than other compounds with Ag3+ ions.
This offers a new example for anionic redox property, a hot topic in the investigation of Li-excess electrode materials. These results demonstrate how unusual structures can be discovered from exceptions in machine learning, which can help us to investigate new physics and novel structural units from existing databases.
Wang is supported by the Director, Office of Science (SC), Basic Energy Science (BES), Materials Science and Engineering Division (MSED), of the US Department of Energy (DOE) under Contract No.DE-AC02-05CH11231 through the Materials Theory program (KC2301). This work is also financially supported by National Key R&D Program of China (2016YFB0700600), Shenzhen Science and Technology Research Grant (ZDSYS201707281026184), and Guangdong Key-lab Project (2017B0303010130).
See the article:
Jianshu Jie, Zongxiang Hu, Guoyu Qian, Mouyi Weng, Shunning Li, Shucheng Li, Mingyu Hu, Dong Chen, Weiji Xiao, Jiaxin Zheng, Lin-Wang Wang and Feng Pan. Discovering unusual structures from exception using big data and machine learning techniques. Science Bulletin, 2019, 64(9)612-616; doi: 10.1016/j.scib.2019.04.015
Feng Pan | EurekAlert!
A new paradigm of material identification based on graph theory
17.06.2019 | Science China Press
Electron beam strengthens recyclable nanocomposite
17.06.2019 | Kanazawa University
The well-known representation of chemical elements is just one example of how objects can be arranged and classified
The periodic table of elements that most chemistry books depict is only one special case. This tabular overview of the chemical elements, which goes back to...
Light can be used not only to measure materials’ properties, but also to change them. Especially interesting are those cases in which the function of a material can be modified, such as its ability to conduct electricity or to store information in its magnetic state. A team led by Andrea Cavalleri from the Max Planck Institute for the Structure and Dynamics of Matter in Hamburg used terahertz frequency light pulses to transform a non-ferroelectric material into a ferroelectric one.
Ferroelectricity is a state in which the constituent lattice “looks” in one specific direction, forming a macroscopic electrical polarisation. The ability to...
Researchers at TU Graz calculate the most accurate gravity field determination of the Earth using 1.16 billion satellite measurements. This yields valuable knowledge for climate research.
The Earth’s gravity fluctuates from place to place. Geodesists use this phenomenon to observe geodynamic and climatological processes. Using...
Discovery by Brazilian and US researchers could change the classification of two species, which appear more akin to jellyfish than was thought.
The tube anemone Isarachnanthus nocturnus is only 15 cm long but has the largest mitochondrial genome of any animal sequenced to date, with 80,923 base pairs....
Researchers at Chalmers University of Technology, Sweden, have discovered a completely new way of capturing, amplifying and linking light to matter at the nanolevel. Using a tiny box, built from stacked atomically thin material, they have succeeded in creating a type of feedback loop in which light and matter become one. The discovery, which was recently published in Nature Nanotechnology, opens up new possibilities in the world of nanophotonics.
Photonics is concerned with various means of using light. Fibre-optic communication is an example of photonics, as is the technology behind photodetectors and...
29.04.2019 | Event News
17.04.2019 | Event News
15.04.2019 | Event News
18.06.2019 | Life Sciences
18.06.2019 | Life Sciences
18.06.2019 | Life Sciences