Barabási-Albert Graph Instances

该数据集包含了随机生成的Barabási-Albert网络实例，用于在约束编程中对值选择启发式算法的训练和评估。在训练过程中，根据实例的大小，考虑了不同的密度因子（从4到15），并在与训练集相同分布的另外20个图上进行评估。数据集按规模分为小型（20-30个节点）、中型（40-50个节点）和大型（80-100个节点）。任务是对组合优化问题的值选择启发式算法进行评估和训练。

This dataset comprises randomly generated Barabási-Albert network instances, intended for training and evaluating value selection heuristics in constraint programming. During the training stage, distinct density factors ranging from 4 to 15 are employed based on the scale of the instances, and evaluation is carried out on an additional 20 graphs adhering to the same distribution as the training set. The dataset is categorized into three size groups: small (20–30 nodes), medium (40–50 nodes), and large (80–100 nodes). The core task of this dataset is to support the evaluation and training of value selection heuristics for combinatorial optimization problems.