当前位置：网站首页>[Code Analysis] Graph small sample anomaly detection method: GDN: Few-shot Network Anomaly Detection via Cross-network Meta-learning

[Code Analysis] Graph small sample anomaly detection method: GDN: Few-shot Network Anomaly Detection via Cross-network Meta-learning

2022-08-08 04:02:00 【vector <>】

Title: Few-shot Network Anomaly Detection via Cross-network Meta-learning
Source code: https://github.com/kaize0409/Meta-GDN_AnomalyDetection
Time: 2021.02
Link: https://arxiv.org/pdf/2102.11165.pdf
Thesis reading notes: [Paper sharing]*** Small sample network anomaly detection method GDN: Few-shot NetworkAnomaly Detection via Cross-network Meta-learning
Directory
I. About abnormality valuator
II. About SGC
III. About data
1. Load data
2. Sample Data
3. Divide data

I. About abnormality valuator

The role of the abnormality valuator is to map the embedding to a scalar score, not as a separate module in the code, but out_feature is set to 1 when calling the model

原网站

版权声明
本文为[vector <>]所创，转载请带上原文链接，感谢
https://yzsam.com/2022/220/202208080353375049.html