智慧的旋风的博客

ogbn_arxiv_GCN_resThis is an improvement of baesline on the ogbn-arxiv dataset.我的代码：https://github.com/ytchx1999/ogbn_arxiv_GCN_resogbn-arxivThe framework is shown in the figure.Improvement Strategy：add skip-connection inspired by DeeperGCNadd ini

PyG-citation-networkGNN模型在引文网络数据集上的代码，包括Cora、Citeseer、Pubmed、ogbn-arxiv我的代码：https://github.com/ytchx1999/PyG-citation-network参考代码：https://github.com/snap-stanford/ogb/blob/master/examples/nodeproppred/arxiv/gnn.pyhttps://github.com/rusty1s/pytorch_ge

GNN-Dataset典型图数据集的加载与使用（基于PyG）。OGB数据集ogbn-arxivogbn-products我的代码：https://github.com/ytchx1999/GNN-Dataset/blob/main/OGBn.ipynbfrom ogb.nodeproppred import PygNodePropPredDataset, Evaluatorimport torch_geometric.transforms as Togbn-arxiv1、加载数

from torch_geometric.datasets import Planetoidimport torch把这里一改就可以实现Citeseer和Pubmed数据集了。dataset_cora = Planetoid(root='./cora/',name='Cora')print(dataset_cora)Cora()import torchimport torch.nn as nnfrom torch_geometric.nn import MessagePassingf

发现PyG已经有了封装好的数据加载、预处理模块了。感觉自己之前处理Cora、Citeseer、Pubmed都白搞了。所以现在我决定站在巨人的肩膀上????，PyG大法好啊！参考资料：https://pytorch-geometric.readthedocs.io/en/latest/notes/introduction.html需要的第三方库torchtorch_geometricfrom torch_geometric.datasets import Planetoidimport to

paper《FAST GRAPH REPRESENTATION LEARNING WITHPYTORCH GEOMETRIC》。文档https://pytorch-geometric.readthedocs.io/en/latest/项目地址https://github.com/rusty1s/pytorch_geometric