An Interoperable Multi Objective Batch Bayesian Optimization Framework for High Throughput Materials Discovery
arxiv(2024)
摘要
In this study, we introduce a groundbreaking framework for materials
discovery, we efficiently navigate a vast phase space of material compositions
by leveraging Batch Bayesian statistics in order to achieve specific
performance objectives. This approach addresses the challenge of identifying
optimal materials from an untenably large array of possibilities in a
reasonable timeframe with high confidence. Crucially, our batchwise methods
align seamlessly with existing material processing infrastructure for
synthesizing and characterizing materials. By applying this framework to a
specific high entropy alloy system, we demonstrate its versatility and
robustness in optimizing properties like strain hardening, hardness, and strain
rate sensitivity. The fact that the Bayesian model is adept in refining and
expanding the property Pareto front highlights its broad applicability across
various materials, including steels, shape memory alloys, ceramics, and
composites. This study advances the field of materials science and sets a new
benchmark for material discovery methodologies. By proving the effectiveness of
Bayesian optimization, we showcase its potential to redefine the landscape of
materials discovery.
更多查看译文
AI 理解论文
溯源树
样例
![](https://originalfileserver.aminer.cn/sys/aminer/pubs/mrt_preview.jpeg)
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要