Skip to content
DraconiticResearch
Search Sources API
194,161 studies31 sourcesUpdated 6 min ago
Peer-reviewedCC BYPLOS One

Machine learning-based thermoelectric materials screening with physical consistency and uncertainty propagation

  • Abid Ahmad1
  • Mohammad Naved Qureshi2
  • Shabbab Ali Algamdi3
  • Shaha Faisal4
4 affiliations
  1. 1Research Center for Eco-environmental Engineering, School of Environment and Civil Engineering, Dongguan University of Technology, Dongguan, China.
  2. 2Electrical Engineering Section, University Polytechnic, Aligarh Muslim University, Aligarh, India.
  3. 3Department of Software Engineering, College of Computer Science and Engineering, Prince Sattam bin Abdulaziz University, Al Kharj, Saudi Arabia.
  4. 4College of Business Administration, Prince Sattam bin Abdulaziz University, Al Kharj, Saudi Arabia.
Read full textPDFDOICite

Abstract

Machine learning (ML) has emerged as a promising strategy for accelerating thermoelectric materials discovery. However, direct prediction of the thermoelectric figure of merit (ZT) often can introduce target leakage and overestimates predictive performance because ZT is intrinsically derived from multiple transport properties. In this work, a physics-guided and uncertainty-aware ML framework is developed for thermoelectric material screening by independently predicting the Seebeck coefficient, electrical conductivity, and thermal conductivity, followed by analytical evaluation of thermoelectric performance. Experimental data from the ESTM database were evaluated using material-wise validation, with exact chemical formulas kept exclusively in either the training or held-out test subset to reduce material-level information leakage. Predictive uncertainty was quantified using ensemble learning and propagated into thermoelectric performance evaluation and uncertainty-aware LCB screening. The proposed framework demonstrates strong predictive capability for individual transport properties while providing physically consistent thermoelectric performance estimation. Although the framework exhibits lower apparent predictive accuracy than direct ZT regression models, it provides improved interpretability and a more physically structured approach to thermoelectric performance evaluation. The results demonstrate that integrating physical consistency, uncertainty propagation, and uncertainty-aware screening provides a robust strategy for data-driven thermoelectric materials screening.

Related

  1. A computed thermoelectric feature database for 50,992 GNoME materials
    Akshay Deshpande
    ChemRxivPreprintLicence not statedGeneral chemistryChemRxiv
  2. Materials design strategies for green thermoelectric generators enabling sustainable and scalable energy harvesting
    Nuur Syahidah Sabran, Iman Aris Fadzallah, Nguyen Van Toan +3
    ISciencePeer-reviewedCC BYFull text1 citedMaterials & polymersEurope PMCCrossref
  3. Nano-Based Advanced Thermoelectric Design: Materials Engineering and Thermal Management Applications
    Jin Wang
    NanomaterialsPeer-reviewedCC BY 4.0Materials & polymersCrossref
  4. Physics-inspired feature engineering and interpretable machine learning for thermoelectric properties of multicomponent materials
    Thi Huynh Ho, Nguyen-Hieu Hoang, Thuat Thanh Trinh
    Materials & DesignPeer-reviewedCC BY5 citedMaterials & polymersOpenAlex
  5. Ensuring Complete Consistency between Hops and Wavefunction Propagation in Fewest-Switches Surface Hopping
    Jonathan R. Mannouch
    arXivPreprintCC BY 4.0Full textPhysical chemistryarXiv

Cite

BibTeXRIS
BibTeX
@article{ahmad2026machine,
  title = {{Machine learning-based thermoelectric materials screening with physical consistency and uncertainty propagation}},
  author = {Abid Ahmad and Mohammad Naved Qureshi and Shabbab Ali Algamdi and Shaha Faisal},
  journal = {PLOS One},
  year = {2026},
  month = nov,
  doi = {10.1371/journal.pone.0360053},
  url = {https://doi.org/10.1371/journal.pone.0360053},
  note = {CC BY},
}
RIS
TY  - JOUR
TI  - Machine learning-based thermoelectric materials screening with physical consistency and uncertainty propagation
AU  - Abid Ahmad
AU  - Mohammad Naved Qureshi
AU  - Shabbab Ali Algamdi
AU  - Shaha Faisal
PY  - 2026
DA  - 2026/11/02
JO  - PLOS One
DO  - 10.1371/journal.pone.0360053
UR  - https://doi.org/10.1371/journal.pone.0360053
AB  - Machine learning (ML) has emerged as a promising strategy for accelerating thermoelectric materials discovery. However, direct prediction of the thermoelectric figure of merit (ZT) often can introduce target leakage and overestimates predictive performance because ZT is intrinsically derived from multiple transport properties. In this work, a physics-guided and uncertainty-aware ML framework is developed for thermoelectric material screening by independently predicting the Seebeck coefficient, electrical conductivity, and thermal conductivity, followed by analytical evaluation of thermoelectric performance. Experimental data from the ESTM database were evaluated using material-wise validation, with exact chemical formulas kept exclusively in either the training or held-out test subset to reduce material-level information leakage. Predictive uncertainty was quantified using ensemble learning and propagated into thermoelectric performance evaluation and uncertainty-aware LCB screening. The proposed framework demonstrates strong predictive capability for individual transport properties while providing physically consistent thermoelectric performance estimation. Although the framework exhibits lower apparent predictive accuracy than direct ZT regression models, it provides improved interpretability and a more physically structured approach to thermoelectric performance evaluation. The results demonstrate that integrating physical consistency, uncertainty propagation, and uncertainty-aware screening provides a robust strategy for data-driven thermoelectric materials screening.
SN  - 1932-6203
ER  -