A Critical Review of Machine Learning Techniques on Thermoelectric Materials

Abstract

Thermoelectric (TE) materials can directly convert heat to electricity and vice versa and have broad application potential for solid-state power generation and refrigeration. Over the past few decades, efforts have been made to develop new TE materials with high performance. However, traditional experiments and simulations are expensive and time-consuming, limiting the development of new materials. Machine learning (ML) has been increasingly applied to study TE materials in recent years. This paper reviews the recent progress in ML-based TE material research. The application of ML in predicting and optimizing the properties of TE materials, including electrical and thermal transport properties and optimization of functional materials with targeted TE properties, is reviewed. Finally, future research directions are discussed.

Machine-learning-assisted discovery of 212-Zintl-phase compounds with ultra-low lattice thermal conductivity

“…Here, six kinds of common supervised learning algorithms which have been successfully applied in the elds of thermal transport and thermoelectrics 55,56 were considered. They are the classical algorithm decision tree (DT), 57 the ensemble learning method random forest (RF), 58 the iterative DT algorithm This journal is © The Royal Society of Chemistry 2024 gradient boosting regressor (GBR), 59 the representative boosting algorithm AdaBoost, 60 the ensemble learning algorithms based on gradient boosting extreme gradient boosting (XGBoost) 61 and light gradient boosting machine (LightGBM).…”

SnSe: The rise of the ultrahigh thermoelectric performance material

“…With the rapid development of computer science and artificial intelligence, materials informatics provides a useful perspective for design and development of new materials based on information of materials. 100,101 For example, machine learning (ML) is a powerful means of exploring new materials and optimal chemical and processing conditions. 102 It has been applied to various research fields such as polymer materials, 103 molecular structure predictions, 104 perovskite structure predictions, 105 and new medicine developments.…”

“…With the rapid development of computer science and artificial intelligence, materials informatics provides a useful perspective for design and development of new materials based on information of materials 100,101 . For example, machine learning (ML) is a powerful means of exploring new materials and optimal chemical and processing conditions 102 .…”

Machine Learning Combined with Weighted Voting Regression and Proactive Searching Progress to Discover ABO3-δ Perovskites with High Oxide Ionic Conductivity

“…Compared to traditional experiments and first-principles calculations, machine learning offers a more efficient and cost-effective way to predict material properties. Currently, machine learning has demonstrated significant success in various fields of materials science, including polymers, alloys, and perovskites. − In particular, the combination of machine learning with experiments or density functional theory (DFT) calculations has emerged as a prominent strategy for the design and discovery of perovskite materials. Hu et al used high-throughput calculations combined with XGboost to model the adsorption energy of two-dimensional halide perovskite A 2 BX 4 with various metal ions.…”