AI Chat Paper
Note: Please note that the following content is generated by AMiner AI. SciOpen does not take any responsibility related to this content.
{{lang === 'zh_CN' ? '文章概述' : 'Summary'}}
{{lang === 'en_US' ? '中' : 'Eng'}}
Chat more with AI
Article Link
Collect
Submit Manuscript
Show Outline
Outline
Show full outline
Hide outline
Outline
Show full outline
Hide outline
Research paper | Open Access

Accurate and explainable machine learning for the power factors of diamond-like thermoelectric materials

Zhe Yanga,1Ye Shenga,1Cong ZhuaJianyue NibZhenyu ZhuaJinyang XiaWu Zhanga,c( )Jiong Yanga( )
Materials Genome Institute, Shanghai University, Shanghai, 200444, China
School of Computer Engineering and Science, Shanghai University, Shanghai, 200444, China
School of Mechanics and Engineering Science, Shanghai University, Shanghai, 200444, China

]]>

Show Author Information

Graphical Abstract

Abstract

The application of machine learning (ML)-based methods to the study of thermoelectric (TE) materials is promising. Although conventional ML algorithms can achieve high prediction performance, their lack of interpretability severely obstructs researchers from extracting material-oriented insights from ML models. In this work, high ML-based prediction performance was achieved with respect to TE power factors (PFs), and the results were well understood by the SHapley Additive exPlanations (SHAP), a method to identify the correlations between targets and descriptors. We designed a robust PF prediction model for diamond-like compounds via a stacking technique, and the model achieved a coefficient of determination value above 0.95 on the test set. From the SHAP analysis, the PFs were negatively correlated with electronegativity and positively correlated with the descriptor "volume per atom" based on the previously reported dataset. TE domain knowledge was adopted to understand these correlations. This work shows that ML models can achieve high accuracy while exhibiting good interpretability, making them useful for materials scientists.

References

[1]

Weishu L, Xiao Y, Gang C, Zhifeng R. Nanomater Energy 2012;1: 42-56.

[2]

Zhu T, Liu Y, Fu C, Heremans J, Snyder J, Zhao X. Adv Mater 2017;29: 1605884.

[3]

Yang J, Xi L, Qiu W, Wu L, Shi X, Chen L, et al. npj Comput Mater 2016;2: 15015.

[4]

Xi L, Pan S, Li X, Xu Y, Ni J, Sun X, et al. J Am Chem Soc 2018;140: 10785-93.

[5]

Graziosi P, Kumarasinghe C, Neophytou N. J Appl Phys 2019;126: 155701.

[6]

Graziosi P, Kumarasinghe C, Neophytou N. ACS Appl Energy Mater 2020;3: 5913-26.

[7]

Liu W, Tan X, Yin K, Liu H, Tang X, Shi J, et al. Phys Rev Lett 2012;108: 166601.

[8]

Zhang J, Liu R, Cheng N, Zhang Y, Yang J, Uher C, et al. Adv Mater 2014;26: 3848-53.

[9]

Pei Y, Shi X, LaLonde A, Wang H, Chen L, Snyder GJ. Nature 2011;473: 66-9.

[10]

Fu C, Zhu T, Liu Y, Xie H, Zhao X. Energ Environ Sci 2015;8: 216-20.

[11]

Chen W, Pöhls J-H, Hautier G, Broberg D, Bajaj S, Aydemir U, et al. J Mater Chem C 2016;4: 4414-26.

[12]

Ricci F, Chen W, Aydemir U, Snyder GJ, Rignanese G-M, Jain A, et al. Sci Data 2017;4: 170085.

[13]

Wang T, Zhang C, Snoussi H, Zhang G. Adv Funct Mater 2020;30: 1906041.

[14]

Xu Y, Xiangmeng W, Li X, Xi L, Ni J, Zhu W, et al. Sci Sin Technol 2019;49: 44-54.

[15]

Wang X, Xu Y, Yang J, Ni J, Zhang W, Zhu W. Comput Mater Sci 2019;169: 109117.

[16]

Wang X, Zeng S, Wang Z, Ni J. J Phys Chem C 2020;124: 8488-95.

[17]

Furmanchuk Ao, Saal JE, Doak JW, Olson GB, Choudhary A, Agrawal A. J Comput Chem 2018;39: 191-202.

[18]

Balachandran PV, Xue D, Theiler J, Hogden J, Lookman T. Sci Rep 2016;6: 19660.

[19]

Bassman L, Rajak P, Kalia RK, Nakano A, Sha F, Sun J, et al. npj Comput Mater 2018;4: 74.

[20]

Sheng Y, Wu Y, Yang J, Lu W, Villars P, Zhang W. npj Comput Mater 2020;6: 171.

[21]

Gladkikh V, Kim DY, Hajibabaei A, Jana A, Myung CW, Kim KS. J Phys Chem C 2020;124: 8905-18.

[22]

Ramprasad R, Batra R, Pilania G, Mannodi-Kanakkithodi A, Kim C. npj Comput Mater 2017;3: 54.

[23]

Schmidt J, Marques MRG, Botti S, Marques MAL. npj Comput Mater 2019;5: 83.

[24]
Lundberg S, Lee S-I. arXiv preprint arXiv 2017. 170507874.
[25]

Ouyang R, Curtarolo S, Ahmetcik E, Scheffler M, Ghiringhelli LM. Phys Rev Mater. 2018. 2: 083802.

[26]

Zhou Z, Cao G, Liu J, Liu H. npj Comput Mater 2020;6: 149.

[27]

Breiman L. Stat Sci 2001;16: 199-231.

[28]
Lundberg SM, Erion GG, Lee S-I. arXiv preprint arXiv 2018;180203888.
[29]

Lundberg SM, Erion G, Chen H, DeGrave A, Prutkin JM, Nair B, et al. Nat Mach Intell 2020;2: 56-67.

[30]
Molnar C, Victoria. British columbia, Canada: lean pub. 2020.
[31]

Wolpert DH. Neural Netw 1992;5: 241-59.

[32]
Ke G, Meng Q, Finley T, Wang T, Chen W, Ma W, et al., editors. Adv neural inf. Process Syst 2017;vol. 30: 3146-54.
[33]
Michailidis MUCL. University College London. 2017.
[34]

Breiman L. Mach learn 2001;45: 5-32.

[35]

Fisher A, Rudin C, Dominici F. J Mach Learn Res 2019;20: 1-81.

[36]

Shapley LS. Contrib Theory Games 1953;2: 307-17.

[37]

Jäger MOJ, Morooka EV, Canova FF, Himanen L, Foster AS. npj Comput Mater 2018;4: 37.

[38]

Park H, Mall R, Ali A, Sanvito S, Bensmail H, El-Mellouhi F. Comput Mater Sci 2020;184: 109858.

[39]

Ward L, Liu R, Krishna A, Hegde VI, Agrawal A, Choudhary A, et al. Phys rev B. 2017. 96: 024104.

[40]

Ward L, Dunn A, Faghaninia A, Zimmermann NER, Bajaj S, Wang Q, et al. Comput Mater Sci 2018;152: 60-9.

[41]

Ong SP, Richards WD, Jain A, Hautier G, Kocher M, Cholia S, et al. Comput Mater Sci 2013;68: 314-9.

[42]

Guyon I, Weston J, Barnhill S, Vapnik V. Mach Learn 2002;46: 389-422.

[43]
Friedman JH. Ann stat. 2001. p. 1189-232.
[44]

Geurts P, Ernst D, Wehenkel L. Mach learn 2006;63: 3-42.

[45]

Smola AJ, Schölkopf B. Stat Comput 2004;14: 199-222.

[46]

Hinton GE. Artif Intell 1989;40: 185-234.

[47]
Murphy K.P. Cambridge, Massachusetts, USA: MIT press; 2012, p. 492-493.
[48]
Rasmussen CE. Summer School on machine learning. 2003. p. 63-71.
Journal of Materiomics
Pages 633-639
Cite this article:
Yang Z, Sheng Y, Zhu C, et al. Accurate and explainable machine learning for the power factors of diamond-like thermoelectric materials. Journal of Materiomics, 2022, 8(3): 633-639. https://doi.org/10.1016/j.jmat.2021.11.010

492

Views

6

Crossref

8

Web of Science

9

Scopus

Altmetrics

Received: 03 June 2021
Revised: 15 November 2021
Accepted: 18 November 2021
Published: 23 November 2021
© 2021 The Chinese Ceramic Society.

This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Return