剪枝,而非重建:为最近邻搜索高效调整α-可达图
Prune, Don't Rebuild: Efficiently Tuning $α$-Reachable Graphs for Nearest Neighbor Search
AI总结:
本文提出RP-Tuning方法,通过剪枝步骤高效调整DiskANN的α参数,无需重建索引,在保持可达性保证的同时提升最近邻搜索效率。
摘要:
向量相似性搜索是现代人工智能和机器学习应用中的基本操作。大多数向量数据库采用基于图的近似最近邻(ANN)搜索算法,如DiskANN(Subramanya等人,2019),这些算法已展现出最先进的实证性能。DiskANN的图构建由一个可达性参数α控制,它在构建时间、查询时间和准确性之间提供权衡。然而,自适应地调整这一权衡通常需要为不同的α值重建索引,这在大规模情况下是不可行的。在本工作中,我们提出了RP-Tuning,一种基于DiskANN剪枝步骤的高效后处理方法,以调整α参数而无需重建整个索引。在先前理论工作的α-可达性框架(Indyk和Xu,2023;Gollapudi等人,2025)内,我们证明了通过RP-Tuning剪枝初始的α-可达图,在一般度量中保持最坏情况可达性保证,并在欧几里得度量中提供改进的保证。实证上,我们显示RP-Tuning在四个公开数据集上加速DiskANN调整,效率提升高达43倍,且开销极小。
英文摘要:
Vector similarity search is an essential primitive in modern AI and ML applications. Most vector databases adopt graph-based approximate nearest neighbor (ANN) search algorithms, such as DiskANN (Subramanya et al., 2019), which have demonstrated state-of-the-art empirical performance. DiskANN's graph construction is governed by a reachability parameter $α$, which gives a trade-off between construction time, query time, and accuracy. However, adaptively tuning this trade-off typically requires rebuilding the index for different $α$ values, which is prohibitive at scale. In this work, we propose RP-Tuning, an efficient post-hoc routine, based on DiskANN's pruning step, to adjust the $α$ parameter without reconstructing the full index. Within the $α$-reachability framework of prior theoretical works (Indyk and Xu, 2023; Gollapudi et al., 2025), we prove that pruning an initially $α$-reachable graph with RP-Tuning preserves worst-case reachability guarantees in general metrics and improved guarantees in Euclidean metrics. Empirically, we show that RP-Tuning accelerates DiskANN tuning on four public datasets by up to $43\times$ with negligible overhead.
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。