HyperAI超神经

Link Property Prediction On Ogbl Citation2

评估指标

Ext. data
Number of params
Test MRR
Validation MRR

评测结果

各个模型在此基准测试上的表现结果

比较表格
模型名称Ext. dataNumber of paramsTest MRRValidation MRR
can-gnns-learn-link-heuristics-a-conciseNo3726740.8891 ± 0.00050.8892 ± 0.0005
open-graph-benchmark-datasets-for-machineNo2811135050.5186 ± 0.04430.5181 ± 0.0436
模型 3No00.5147 ± 0.00000.5119 ± 0.0000
graphgpt-graph-learning-with-generative-preNo467841280.9055 ± 0.00160.9042 ± 0.0014
inductive-representation-learning-on-largeNo4602890.8044 ± 0.00100.8054 ± 0.0009
inductive-representation-learning-on-largeNo4602890.8260 ± 0.00360.8263 ± 0.0033
adaptive-graph-diffusion-networks-with-hopNo3067160.8549 ± 0.00290.8556 ± 0.0033
network-in-graph-neural-networkNo11344020.8891 ± 0.00220.8879 ± 0.0022
node2vec-scalable-feature-learning-forNo3749111050.6141 ± 0.00110.6124 ± 0.0011
circle-feature-graphormer-can-circle-featuresNo6862530.8997 ± 0.00150.8987 ± 0.0011
algorithm-and-system-co-design-for-efficientNo796170.8883 ± 0.00180.8891 ± 0.0021
pure-message-passing-can-estimate-commonNo7497572830.9072 ± 0.00120.9074 ± 0.0011
simplifying-subgraph-representation-learningNo1422750010.8814 ± 0.00080.8809 ± 0.0074
graphgpt-graph-learning-with-generative-preNo1330968320.9305 ± 0.00200.9295 ± 0.0022
graphsaint-graph-sampling-based-inductiveNo2964490.7985 ± 0.00400.7975 ± 0.0039
模型 16No7495585280.8432 ± 0.00030.8422 ± 0.0002
semi-supervised-classification-with-graphNo2964490.8474 ± 0.00210.8479 ± 0.0023
cluster-gcn-an-efficient-algorithm-forNo2964490.8004 ± 0.00250.7994 ± 0.0025
模型 19No2568020.8957 ± 0.00100.8948 ± 0.0008
模型 20No00.5189 ± 0.00000.5167 ± 0.0000
pairwise-learning-for-neural-link-predictionNo1465145510.8492 ± 0.00290.8490 ± 0.0031
模型 22No1665310.8796 ± 0.00080.8793 ± 0.0008
revisiting-graph-neural-networks-for-link-1No2608020.8767 ± 0.00320.8757 ± 0.0031