Musk v1

麝香数据集描述了一组分子，目标是从非麝香中检测出麝香。该数据集描述了一组 92 个分子，其中 47 个被人类专家判断为麝香，其余 45 个分子被判断为非麝香。有 166 个可用的特征可以根据分子的形状来描述分子。

The Musk Dataset comprises a set of molecules, with the core task of detecting musk molecules from non-musk ones. This dataset contains 92 molecules in total: 47 of them are labeled as musk molecules by human experts, while the remaining 45 are categorized as non-musk molecules. A total of 166 features are available to characterize molecules based on their molecular shapes.

搜集汇总

背景与挑战

背景概述

Musk v1数据集包含92个分子样本（47个麝香，45个非麝香），提供166个分子形状特征，用于麝香检测任务。数据集由俄勒冈州立大学于1994年发布，适用于分类算法研究。