Diversity-Aware Active Learning Builds Materials Datasets for Current and Future Objectives
Abstract
Materials science data collection can be expensive, making dataset reuse and long-term utility important for future discovery. In practice, data collection often prioritizes properties of current interest while other properties are excluded from the acquisition process. This can produce datasets that are well suited for current objectives but lose information needed for properties of future interest. Here, we present how to construct datasets that remain informative for both. We develop and evaluate acquisition frameworks that combine predictive disagreement with diversity in material features, predicted outcomes, or both. We test these approaches across four DFT datasets and an experimental thermoelectric dataset. Diversity inclusion in dataset construction can reduce present and future objectives trade-off by preserving a broader range of materials and property behavior during acquisition. Across all datasets, the Non-dominated Sorting Genetic Algorithm II with disagreements and feature diversity (NSGA-II Dis. + diversity) framework developed in this work provides the most robust balance between current and future objectives. It maintains prediction performance above random sampling for current objectives while mitigating bias on excluded properties and, in many cases, improving their performance beyond random sampling. This diversity-aware acquisition framework provides a practical route to building materials datasets with unbiased quality entries and mitigating cold-start limitations in subsequent modeling and discovery campaigns.
Related
- Different representation learning objectives recover distinct latent structures from the same psychometric data
- Supervised Learning Objectives for High-Dimensional Treatment Effects
- Deep Reinforcement Learning with Buffered Quantile Objectives
- Constrained Reinforcement Learning-Based Framework for DC Microgrid: Current Sharing and Voltage Regulation
- Machine learning-based thermoelectric materials screening with physical consistency and uncertainty propagation
BibTeX
@misc{castaneda2026diversityaware,
title = {{Diversity-Aware Active Learning Builds Materials Datasets for Current and Future Objectives}},
author = {Rafael Espinosa Castañeda and Ashley Dale and Hongchen Wang and Yonatan Kurniawan and Hao Wan and Runze Zhang and Adji Bousso Dieng and Kangming Li and Jason Hattrick-Simpers},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2605.05104},
eprint = {2605.05104},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2605.05104},
}
RIS
TY - UNPB
TI - Diversity-Aware Active Learning Builds Materials Datasets for Current and Future Objectives
AU - Rafael Espinosa Castañeda
AU - Ashley Dale
AU - Hongchen Wang
AU - Yonatan Kurniawan
AU - Hao Wan
AU - Runze Zhang
AU - Adji Bousso Dieng
AU - Kangming Li
AU - Jason Hattrick-Simpers
PY - 2026
DA - 2026/10/17
PB - arXiv
DO - 10.48550/arxiv.2605.05104
UR - https://arxiv.org/abs/2605.05104
AB - Materials science data collection can be expensive, making dataset reuse and long-term utility important for future discovery. In practice, data collection often prioritizes properties of current interest while other properties are excluded from the acquisition process. This can produce datasets that are well suited for current objectives but lose information needed for properties of future interest. Here, we present how to construct datasets that remain informative for both. We develop and evaluate acquisition frameworks that combine predictive disagreement with diversity in material features, predicted outcomes, or both. We test these approaches across four DFT datasets and an experimental thermoelectric dataset. Diversity inclusion in dataset construction can reduce present and future objectives trade-off by preserving a broader range of materials and property behavior during acquisition. Across all datasets, the Non-dominated Sorting Genetic Algorithm II with disagreements and feature diversity (NSGA-II Dis. + diversity) framework developed in this work provides the most robust balance between current and future objectives. It maintains prediction performance above random sampling for current objectives while mitigating bias on excluded properties and, in many cases, improving their performance beyond random sampling. This diversity-aware acquisition framework provides a practical route to building materials datasets with unbiased quality entries and mitigating cold-start limitations in subsequent modeling and discovery campaigns.
ER -