baseline

该预印本系统比较了在不同推理信息假设（无化学式、已知化学式、峰值带子式注释）下的最近邻基线方法，用于从MS/MS光谱预测分子指纹，评估了五种谱相似度在NPLIB1与MassSpecGym数据集上的表现。

This paper proposes using random feature selection as a baseline and empirically shows that many state-of-the-art unsupervised feature selection methods are outperformed by random selection in both performance and efficiency.