合成图数据集
收藏资源简介:
合成图数据集由悉尼大学数学与统计学院创建,用于链接预测算法的基准测试。该数据集包含10个合成图,每个图具有微观尺度的图案和介观尺度的社区结构,这些是复杂网络的普遍特征。数据集的创建过程结合了理论分析和随机图模型,旨在生成具有可预测性的图结构。该数据集主要应用于链接预测领域,旨在评估和改进现有的链接预测方法,解决在复杂网络中预测缺失链接的问题。
This synthetic graph dataset was created by the School of Mathematics and Statistics, the University of Sydney, for benchmarking link prediction algorithms. This dataset contains 10 synthetic graphs, each featuring micro-scale patterns and meso-scale community structures, which are universal characteristics of complex networks. The dataset was developed by combining theoretical analysis and random graph models, with the goal of generating graph structures with predictable properties. This dataset is primarily applied in the field of link prediction, aiming to evaluate and improve existing link prediction methods and address the problem of predicting missing links in complex networks.

- 1Synthetic graphs for link prediction benchmarking悉尼大学数学与统计学院 · 2024年



