DOE PAGES title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Overcoming sparse datasets with multi-task learning as applied to high entropy alloys

Journal Article · · Machine Learning: Science and Technology

Abstract The design of novel High Entropy Alloys for use in high-temperature applications is an area of active interest due to their potential to provide exceptional properties compared to conventional alloys. Since the increased popularity of machine learning, an important cog in the design process has been training surrogate models on alloy properties. However, these Single-Task models are trained on individual mechanical properties and do not take advantage of the relatedness between properties. Multi-Task models can capture the interdependencies between tasks, leading to potentially more accurate predictions for all tasks. In this paper, we investigate if Multi-Task models can show improvement over Single-Task models when used for predicting the mechanical properties of these alloys. To ensure fair evaluation between the models, we apply L 0 regularization and skip connections to the models, which allows them to adjust the number of model parameters and depth for optimal performance. We find that the Multi-Task models can leverage task relationships to perform better than Single-Task models, especially for high amounts of missing data in the tasks. Furthermore, adding simple auxiliary targets can boost Multi-Task performance even further despite not being effective as input descriptors to single-task models themselves. We anticipate that the proposed strategies can achieve more accurate predictions and consequently enable better design capabilities for such data-constrained domains without incurring much additional computational cost.

Sponsoring Organization:
USDOE; USDOE Advanced Research Projects Agency - Energy (ARPA-E)
Grant/Contract Number:
AR0001435
OSTI ID:
2520192
Journal Information:
Machine Learning: Science and Technology, Journal Name: Machine Learning: Science and Technology Journal Issue: 1 Vol. 6; ISSN 2632-2153
Publisher:
IOP PublishingCopyright Statement
Country of Publication:
United Kingdom
Language:
English

References (32)

Nanostructured High-Entropy Alloys with Multiple Principal Elements: Novel Alloy Design Concepts and Outcomes journal May 2004
Zur Theorie der Elastizit�tsgrenze und der Festigkeit kristallinischer K�rper journal July 1926
Machine learning assisted design of high entropy alloys with desired property journal May 2019
A machine learning-based alloy design system to facilitate the rational design of high entropy alloys with enhanced hardness journal January 2022
Multi-objective materials bayesian optimization with active learning of design constraints: Design of ductile refractory multi-principal-element alloys journal September 2022
Extreme sparsification of physics-augmented neural networks for interpretable model discovery in mechanics journal June 2024
Investigating representation schemes for surrogate modeling of High Entropy Alloys journal October 2023
Miedema Calculator: A thermodynamic platform for predicting formation enthalpies of alloys within framework of Miedema’s Theory journal December 2016
Comprehensive data compilation on the mechanical properties of refractory high-entropy alloys journal December 2018
Machine learning – informed development of high entropy alloys with enhanced corrosion resistance journal February 2024
Investigation of phase-transformation path in TiZrHf(VNbTa)x refractory high-entropy alloys and its effect on mechanical property journal December 2021
Data-driven design of high bulk modulus high entropy alloys using machine learning journal December 2024
Machine learning assisted design of high-entropy alloys with ultra-high microhardness and unexpected low density journal February 2024
Predicting and understanding the ductility of BCC high entropy alloys via knowledge-integrated machine learning journal March 2024
Machine learning-assisted design of refractory high-entropy alloys with targeted yield strength and fracture strain journal October 2024
Microstructural development in equiatomic multicomponent alloys journal July 2004
Mechanical properties of pure elements from a comprehensive first-principles study to data-driven insights journal December 2024
Yield strength prediction of high-entropy alloys using machine learning journal March 2021
Polymer informatics with multi-task learning journal April 2021
Composition design of high-entropy alloys with deep sets learning journal April 2022
Element-wise representations with ECNet for material property prediction and applications in high-entropy alloys journal December 2022
A neural network model for high entropy alloy design journal April 2023
On exploiting nonparametric kernel-based probabilistic machine learning over the large compositional space of high entropy alloys for optimal nanoscale ballistics journal July 2024
Machine learning-based inverse design for single-phase high entropy alloys journal October 2022
Explainable machine learning assisted molecular-level insights for enhanced specific stiffness exploiting the large compositional space of AlCoCrFeNi high entropy alloys journal June 2024
Machine learning–enabled high-entropy alloy discovery journal October 2022
When Federated Learning Meets Privacy-Preserving Computation journal October 2024
Research Progress of Refractory High Entropy Alloys: A Review journal November 2022
Development and exploration of refractory high entropy alloys—A review journal June 2018
nimCSO: A Nim package for Compositional Space Optimization journal November 2024
Machine Learning-Based Strength Prediction for Refractory High-Entropy Alloys of the Al-Cr-Nb-Ti-V-Zr System journal November 2021
Code and data for 'Overcoming sparse datasets with Multi-Task Learning as applied to High Entropy Alloys' software January 2025