arXiv：cs.LG· Mario Koddenbrock, Christoph Lange, Robin Legner, Martin J\"ager, Martin K\"ogler, Mariano N. Cruz Bournazou, Peter Neubauer, Felix Biessmann, Erik Rodner·· 4 小时前AI 评分42
RamanBench：面向拉曼光谱机器学习的大规模基准
RamanBench: A Large-Scale Benchmark for Machine Learning on Raman Spectroscopy
AI 导读
RamanBench 是首个面向拉曼光谱机器学习的大规模、完全可复现基准，统一 74 个数据集（含 16 个首次发布）、325,668 条光谱，覆盖四个领域的分类与回归任务。基准在标准化协议下评测 28 个模型，涵盖 PLS、RamanNet、TabPFN 等经典、领域专用、表格基础模型与时序方法，结果显示 TFM 稳定优于领域专用和梯度提升基线。但没有任何方法能跨数据集泛化，暴露出根本性缺口。
正文
Abstract:Machine Learning (ML) has transformed many scientific fields, yet key applications still lack standardized benchmarks. Raman spectroscopy, a widely used technique for non-invasive molecular analysis, is one such field where progress is limited by fragmented datasets, inconsistent evaluation, and models that fail to capture the structure of spectral data. We introduce RamanBench, the first large-scale, fully reproducible benchmark for ML on Raman spectroscopy, consisting of streamlined data access, evaluation protocols and code, as well as a live leaderboard. It unifies 74 datasets (including 16 first released with this benchmark) across four domains, comprising 325,668 spectra and spanning classification and regression tasks under diverse experimental conditions. We benchmark 28 models under a standardized protocol, including classical methods (e.g., PLS), Raman-specific (e.g., RamanNet), Tabular Foundation Model (TFM) (e.g., TabPFN), and time-series approaches (e.g., ROCKET). TFM consistently outperform domain-specific and gradient boosting baselines, while time-series models remain competitive. However, no method generalizes across datasets, revealing a fundamental gap. Therefore, we invite the community to contribute new approaches to our living benchmark, with the potential to accelerate advances in critical applications such as medical diagnostics, biological research, and materials science.
| Comments: | this https URL |
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2605.02003 [cs.LG] |
| (or arXiv:2605.02003v3 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2605.02003 arXiv-issued DOI via DataCite |
|
| Journal reference: | NeurIPS 2026 |
Submission history
From: Mario Koddenbrock [view email]
[v1]
Sun, 3 May 2026 18:12:42 UTC (4,736 KB)
[v2]
Wed, 6 May 2026 11:03:18 UTC (4,680 KB)
[v3]
Wed, 7 Oct 2026 12:10:59 UTC (4,720 KB)
来源：arXiv：cs.LG · arxiv.org