使用扩散模型进行引导轨迹生成以用于离线基于模型的优化
Guided Trajectory Generation with Diffusion Models for Offline Model-based Optimization
AI总结：
针对离线 MBO 中正向/逆向方法的保守与多模态学习难题，本文提出 GTG：构建带局部性偏置的高分合成轨迹，训练条件扩散模型并引导采样，再用代理函数筛选高保真设计，在 Design-Bench 上优于竞争基线。
摘要：
优化复杂的高维黑盒函数在科学和工程领域无处不在。遗憾的是，在大多数情况下，由于时间和安全约束，这些函数的在线评估受到限制。在离线基于模型的优化（MBO）中，我们的目标是仅使用预先存在的离线数据集找到一个使目标函数最大化的设计。尽管先前的方法考虑使用正向或逆向方法来解决该问题，但这些方法受到保守性以及学习高度多模态映射的困难所限制。最近，出现了一种新兴范式，即利用从离线数据集构建的合成轨迹来学习改进解决方案。在本文中，我们提出了一种新颖的条件生成建模方法，以生成通向高分区域的轨迹。首先，我们使用数据集构建通向高分区域的合成轨迹，同时注入局部性偏置以保持一致的改进方向。然后，我们训练一个条件扩散模型，以分数为条件生成轨迹。最后，我们在引导下从训练好的模型中采样多条轨迹，以探索数据集之外的高分区域，并使用代理函数在生成的轨迹中选择高保真设计。大量实验结果表明，我们的方法在 Design-Bench 及其实际变体上优于具有竞争力的基线。代码已公开发布于 https://github.com/dbsxodud-11/GTG。
英文摘要：
Optimizing complex and high-dimensional black-box functions is ubiquitous in science and engineering fields. Unfortunately, the online evaluation of these functions is restricted due to time and safety constraints in most cases. In offline model-based optimization (MBO), we aim to find a design that maximizes the target function using only a pre-existing offline dataset. While prior methods consider forward or inverse approaches to address the problem, these approaches are limited by conservatism and the difficulty of learning highly multi-modal mappings. Recently, there has been an emerging paradigm of learning to improve solutions with synthetic trajectories constructed from the offline dataset. In this paper, we introduce a novel conditional generative modeling approach to produce trajectories toward high-scoring regions. First, we construct synthetic trajectories toward high-scoring regions using the dataset while injecting locality bias for consistent improvement directions. Then, we train a conditional diffusion model to generate trajectories conditioned on their scores. Lastly, we sample multiple trajectories from the trained model with guidance to explore high-scoring regions beyond the dataset and select high-fidelity designs among generated trajectories with the proxy function. Extensive experiment results demonstrate that our method outperforms competitive baselines on Design-Bench and its practical variants. The code is publicly available in \texttt{https://github.com/dbsxodud-11/GTG}.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
