用细粒度引导实现一次优化通用化，提升效率与泛化能力
Make Optimization Once and for All with Fine-grained Guidance
- 通过全局采样增强而非局部更新，构建通用优化学习框架
- 仅需分钟级训练即达竞品数小时效果，样本多样性显著提升性能
- 适用于多种任务，尤其适合追求高效泛化的优化场景
学习优化（L2O）通过集成神经网络提升优化效率，在重拟优化器、迭代生成新解或直接生成解方面表现优异。然而传统L2O方法依赖特定优化流程，设计复杂，限制了可扩展性与泛化能力。本文提出通用学习优化框架Diff-L2O，聚焦从更广视角增强采样解的多样性，而非仅关注实际优化过程中的局部更新。同时给出相关泛化界，表明Diff-L2O的样本多样性带来更好性能，该边界可推广至其他领域，讨论多样性、均值-方差关系及不同任务。实验验证，Diff-L2O仅需分钟级训练即可媲美其他需小时级训练的方法，展现强兼容性。
原文摘要 · Abstract (English)
Learning to Optimize (L2O) enhances optimization efficiency with integrated neural networks. L2O paradigms achieve great outcomes, e.g., refitting optimizer, generating unseen solutions iteratively or directly. However, conventional L2O methods require intricate design and rely on specific optimization processes, limiting scalability and generalization. Our analyses explore general framework for learning optimization, called Diff-L2O, focusing on augmenting sampled solutions from a wider view rather than local updates in real optimization process only. Meanwhile, we give the related generalization bound, showing that the sample diversity of Diff-L2O brings better performance. This bound can be simply applied to other fields, discussing diversity, mean-variance, and different tasks. Diff-L2O's strong compatibility is empirically verified with only minute-level training, comparing with other hour-levels.
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。