arXiv：cs.LG· Ling Min Serena Khoo·· 3 小时前AI 评分24
不同假设下 MS/MS 谱图分子指纹预测的最近邻基线对比
Nearest-neighbour baselines for fingerprint prediction from MS/MS spectra under different assumptions
AI 导读
最新研究表明，最近邻检索为 MS/MS 谱图分子指纹预测提供了强基线，部分变体已匹配或超越当前深度学习模型。该报告系统比较了多种最近邻变体，揭示推理阶段可用信息假设的差异如何影响性能，目标是建立更严格的基线以支持更严谨的基准测试并更好衡量该领域进展。
正文
Abstract:It has recently been shown that nearest-neighbour retrieval provides a strong baseline for molecular fingerprint prediction from MS/MS spectra, with several variants matching or outperforming current deep learning models (Khoo and Barzilay, 2026; Liu et al., 2026; Gupta et al., 2026). Importantly, "nearest neighbour" encompasses a family of retrieval methods that differ in the information assumed to be available at inference. In this report, we systematically compare several nearest-neighbour variants and show how these differing assumptions affect performance. Our goal is to establish stricter baselines that enable more rigorous benchmarking and better measure progress in this area.
| Comments: | 6 pages, 2 figures |
| Subjects: | Machine Learning (cs.LG); Computational Engineering, Finance, and Science (cs.CE); Machine Learning (stat.ML) |
| Cite as: | arXiv:2610.02249 [cs.LG] |
| (or arXiv:2610.02249v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2610.02249 arXiv-issued DOI via DataCite |
Submission history
From: Ling Min Serena Khoo [view email]
[v1]
Wed, 30 Sep 2026 18:41:00 UTC (12 KB)
来源：arXiv：cs.LG · arxiv.org