Multi-task deep-learning optimization of trade-off properties for superior-performance Fe-based soft magnetic alloys

Abstract Fe-based amorphous alloys are promising soft magnetic materials for developing next-generation devices with high frequency and efficiency. However, optimizing Fe-based alloys with ultra-high saturation magnetic flux density ( B s ), ultra-low coercivity ( H c ), and good glass-forming ability remains a notorious challenge owing to the vast composition space and complex trade-offs among these properties. Thus, conventional design methods face great challenges. Here, we develop a generative multi-task deep learning (GMTDL) approach to achieve simultaneous optimization of compositions and trade-off properties. The GMTDL can sufficiently exploit and share knowledge from datasets across different tasks, despite the limitations and imbalances of these datasets. Therefore, it exhibits superior performance in predicting alloys with multiple targeted properties, outperforming previous machine learning-based design strategies. Moreover, the GMTDL can also tailor compositions, providing an efficient way to regulate properties and generate desired candidates for further experimental processing. The validity and reliability of GMTDL are rigorously tested by benchmarking against Fe-based alloys reported very recently. Moreover, some new alloys with ultra-high B s and ultra-low H c are predicted. The optimal content windows of key elements and their synergistic effects are also unraveled, providing practical guidance. Thus, our study establishes an effective and reliable paradigm for simultaneous prediction and optimization of high-performance materials with multiple properties.