ASlib

该数据集名为ASlib，是一个精心整理的集合，包含了超过25个不同的算法选择问题，这些问题被称为“场景”，基于不同的算法问题类别，如SAT、TSP、CSP等。每个场景包含了若干实例，这些实例下，一组算法的性能已经通过设定一定的截止时间来避免过长的算法运行时间而被评估。此外，该数据集中某些实例存在缺失的特征值，这些值是通过截至当前点所有实例特征值的平均值来填充的，同时特征值也被缩放为单位向量。任务方面，该数据集针对的是算法选择问题。

The dataset is named ASlib, which is a meticulously curated collection encompassing over 25 distinct algorithm selection problems, termed "scenarios", based on diverse categories of algorithmic problems such as SAT, TSP, CSP, and others. Each scenario includes multiple instances, where the performance of a suite of algorithms has been evaluated by imposing a runtime cutoff to avoid excessively long algorithm execution durations. Additionally, some instances in this dataset contain missing feature values, which are imputed using the mean of all feature values of the instances up to the current time point, and the feature values are also scaled to unit vectors. The task targeted by this dataset is algorithm selection.