ShanghaiTech University Knowledge Management System
| DiOpt: Self-supervised Diffusion for Constrained Optimization | |
| 2025-02-14 | |
| 状态 | 已发表 |
| 摘要 | Recent advances in diffusion models show promising potential for learning-based optimization by leveraging their multimodal sampling capability to escape local optima. However, existing diffusion-based optimization approaches, often reliant on supervised training, lacks a mechanism to ensure strict constraint satisfaction which is often required in real-world applications. One resulting observation is the distributional misalignment, i.e. the generated solution distribution often exhibits small overlap with the feasible domain. In this paper, we propose DiOpt, a novel diffusion paradigm that systematically learns near-optimal feasible solution distributions through iterative self-training. Our framework introduces several key innovations: a target distribution specifically designed to maximize overlap with the constrained solution manifold; a bootstrapped self-training mechanism that adaptively weights candidate solutions based on the severity of constraint violations and optimality gaps; and a dynamic memory buffer that accelerates convergence by retaining high-quality solutions over training iterations. To our knowledge, DiOpt represents the first successful integration of self-supervised diffusion with hard constraint satisfaction. Evaluations on diverse tasks, including power grid control, motion retargeting, wireless allocation demonstrate its superiority in terms of both optimality and constraint satisfaction. |
| 语种 | 英语 |
| DOI | arXiv:2502.10330 |
| 相关网址 | 查看原文 |
| 出处 | Arxiv |
| 收录类别 | PPRN.PPRN |
| WOS记录号 | PPRN:121679931 |
| WOS类目 | Computer Science, Artificial Intelligence |
| 文献类型 | 预印本 |
| 条目标识符 | https://kms.shanghaitech.edu.cn/handle/2MSLDSTB/514095 |
| 专题 | 信息科学与技术学院_博士生 信息科学与技术学院_本科生 信息科学与技术学院_PI研究组_石野组 |
| 通讯作者 | Shi, Ye |
| 作者单位 | 1.ShanghaiTech Univ, Shanghai, Peoples R China 2.MoE, Key Lab Intelligent Percept & Human Machine Collaborat, Shanghai, Peoples R China 3.China Mobile Commun Co Ltd Res Inst, Beijing, Peoples R China 4.Shanghai Jiao Tong Univ, Shanghai, Peoples R China 5.Chinese Univ Hong Kong Shenzhen, Shenzhen, Peoples R China |
| 推荐引用方式 GB/T 7714 | Ding, Shutong,Zhou, Yimiao,Hu, Ke,et al. DiOpt: Self-supervised Diffusion for Constrained Optimization. 2025. |
| 条目包含的文件 | ||||||
| 条目无相关文件。 | ||||||
修改评论
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。