多任务学习在合金性能预测中效果两极分化，需按任务类型选择模型策略。
Multi-Task Learning for Metal Alloy Property Prediction: An Empirical Study of Negative Transfer and Mitigation Strategies
- 基于5.4万条数据，发现不同性能预测任务间存在严重梯度冲突。
- 电阻率和硬度的回归性能因多任务学习显著下降，但非晶形成分类召回率提升。
- 提出分场景策略：高精度用单任务模型，高通量筛选用多任务模型。
材料科学中的多任务学习依赖于物理相关属性共享可学习表征的假设。我们基于包含54,028个样本的金属合金数据集，挑战这一假设。结果揭示显著分歧：多任务学习显著降低电阻率和硬度的回归性能，但提升非晶形成能力分类的召回率。我们发现根源在于功能形式不匹配——如电阻率呈多项式依赖，而硬度涉及复杂相互作用——导致优化过程中梯度严重错位。评估深度不平衡回归技术后，发现投影冲突梯度（PCGrad）可恢复单任务性能，而标签分布平滑结合梯度归一化实现最佳整体平衡。因此，我们提出战略框架：高精度表征使用独立模型，高通量筛选则采用多任务学习以最大化召回率。这些发现支持‘材料性能聚类’假说，表明不同物理机制需专用优化策略以克服负迁移问题。
原文摘要 · Abstract (English)
Multi-task learning (MTL) in materials science relies on the assumption that physically related properties share learnable representations. We challenge this assumption using a 54,028-sample metal alloy dataset exhibiting extreme task-level imbalance. Our results reveal a striking dichotomy: MTL significantly degrades regression performance for resistivity and hardness but improves classification recall for amorphous-forming ability. We trace this divergence to mismatched functional forms--such as resistivity's polynomial dependence versus hardness's complex interactions--which cause severe gradient misalignment during optimization. Evaluating Deep Imbalanced Regression techniques, we find that projecting conflicting gradients (PCGrad) recovers single-task performance, while combining label distribution smoothing with gradient normalization achieves the best overall balance. Consequently, we propose a strategic framework: utilize independent models for high-precision characterization, but employ MTL for high-throughput screening where recall is paramount. These findings support a "materials property clustering" hypothesis, suggesting that distinct physical mechanisms require specialized optimization strategies to overcome negative transfer.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。