search

TrustAGI-Lab / CoLA

[TNNLS] Anomaly Detection on Attributed Networks via Contrastive Self-Supervised Learning
MIT License
81 stars 20 forks source link

implementation details #3

Open LiveTowardTheSea opened 2 years ago

LiveTowardTheSea commented 2 years ago

in run.py, line 261, why the added_feat_zero_row is in the middle of node sequence, according to the implementation of getting adj in a batch, this zero should be appended to th end of each sample. Could somebody explain it to me?

yesy6sysu commented 2 months ago

Actually it is the last index in the subgraph generated by rwr, which is the target node according to the code subv[i].append(i). But I don't understand why an extra node is added before the input.