Artificial Intelligence and Machine

Learning techniques for novel

Material Design and Discovery
Salman Zafar (28766)

MAT 68001 - Special Topics in MAT Foundations of Nano Engineering

Material Design and Discovery - Introduction

Lots of material existed in present day world – 530

Aluminium alloys and 22000 grades of steel.
14 grand challenges of 21st century requires
discovery of new advanced materials.

How do scientist discover materials?

 Artificial Intelligence and Machine Learning - Basics

Radiation Oncology in the Era of Big Data and Machine Learning for Precision Medicine - Scientific Figure on ResearchGate. Available from:
https://www.researchgate.net/figure/A-generic-machine-learning-workflow_fig3_330542014 [accessed 18 Aug, 2021]
 Artificial Intelligence and Machine Learning - Basics

Guo K, Yang Z, Yu C-H, Buehler MJ. Artificial intelligence and machine learning in design of mechanical materials. Mater Horizons 2021;8:1153–72.
https://doi.org/10.1039/D0MH01451F
 Prediction of Super Hard Materials
ML Model and Data Extraction: Support
Vector Regression (SVR) machine learning
model was used and 3288 elastic moduli (B
and G) were used to training and validation
of ML model.

Descriptors: 34 distinct compositional

variables were used such as position in
periodic table, electronic structure, and
physical properties. 14 descriptors were also
used to crystal system and unit cell volume.

Results: WB4, TaC, WC, ReB2 were

predicated super hard materials that are
already existed, hence confirming the
accuracy of the model.

Tehrani AM, Oliynyk AO, Parry M, Rizvi Z, Couper S, Lin F, et al. Machine Learning Directed Search for Ultraincompressible, Superhard Materials 2018.
https://doi.org/10.1021/jacs.8b02717
 Designing Graphene Nanopores for Water Desalination

AI framework

Trained and Predicted Water flux and Ion rejection rate

Wang Y, Cao Z, Farimani AB. Efficient water desalination with graphene nanopores obtained using artificial intelligence n.d. https://doi.org/10.1038/s41699-021-00246-9.
Prediction of Protein Structure - Alphafold

Formation of 3D protein structure

AIphafold framework
 Challenges in using AI/ML

Zhou Q, Lu S, Wu Y, Wang J. Property-Oriented Material Design Based on a Data-Driven Machine Learning Technique. J Phys Chem Lett 2020;11:3920–7.
https://doi.org/10.1021/ACS.JPCLETT.0C00665.
Thank you
References
Tehrani AM, Oliynyk AO, Parry M, Rizvi Z, Couper S, Lin F, et al. Machine Learning Directed Search for Ultraincompressible, Superhard Materials 2018.
https://doi.org/10.1021/jacs.8b02717.

Guo K, Yang Z, Yu C-H, Buehler MJ. Artificial intelligence and machine learning in design of mechanical materials. Mater Horizons 2021;8:1153–72.
https://doi.org/10.1039/D0MH01451F.

Wang Y, Cao Z, Farimani AB. ARTICLE Efficient water desalination with graphene nanopores obtained using artificial intelligence n.d.
https://doi.org/10.1038/s41699-021-00246-9.

Jumper J, Evans R, Pritzel A, Green T, Figurnov M, Ronneberger O, et al. Highly accurate protein structure prediction with AlphaFold. Nat 2021 2021:1–11.
https://doi.org/10.1038/s41586-021-03819-2.

Zhou Q, Lu S, Wu Y, Wang J. Property-Oriented Material Design Based on a Data-Driven Machine Learning Technique. J Phys Chem Lett 2020;11:3920–7.
https://doi.org/10.1021/ACS.JPCLETT.0C00665.

Mazhnik E, Oganov AR. Application of machine learning methods for predicting new superhard materials. J Appl Phys 2020;128:75102.
https://doi.org/10.1063/5.0012055.

Jiazhen Cai, Xuan Chu, Kun Xu, Hongbo Li, Jing Wei. Machine learning-driven new material discovery. Nanoscale Adv 2020;2:3115–30.
https://doi.org/10.1039/D0NA00388C.

Wang H, Ji Y, Li Y. Simulation and design of energy materials accelerated by machine learning. Wiley Interdiscip Rev Comput Mol Sci 2020;10:e1421.
https://doi.org/10.1002/WCMS.1421.

Cutler J, Dickenson M. Introduction to Machine Learning with Python. 2020. https://doi.org/10.1007/978-3-030-36826-5_10.

Brownlee J. 00 ML Mastery - Understand You Data, Create Accurate Models and Work Projects End-to-End. 2017;91:399–404.