一是不同传感器之间有着非常不同的行为,即图中节点的数值和分布差异很大,因此需要考虑如何对传感器,即图中节点进行特征表示;二是GNNs的输入必须是整个图,即包括图中节点的特征表示以及各节点的连接关系,而在本文场景中,各节点之间的关系都是未知的(以往的方法是直接认为各节点之间都存在关系,即使用完全图表征各节点...
论文链接:[2106.06947v1] Graph Neural Network-Based Anomaly Detection in Multivariate Time Series (arxiv.org) 主要内容 论文提出了一种图偏差网络(GDN)框架用于多元时间序列异常检测,该框架可以实验对一个系统中各个传感器之间的结构关系的建模,传感器异常检测以及异常传感器的定位等,通过两个水处理厂的数据集进行...
Graph neural network-based anomaly detection for human cyber physical systemsChengwen XueLimei LinYanze HuangXiaoding Wang
Xiao et al. [11] proposed a graph embedding approach to perform anomaly detection on network flows. The authors first converted the network flows into a first-order and secondorder graph. The first-order graph learns the latent features from the perspective of a single host by using its IP ...
smoothedScore = movmean(scorePerTime,windowSize);end References [1] A. Deng and B. Hooi, “Graph neural network-based anomaly detection in multivariate time series,” in Proceedings of the 35th AAAI Conference on Artificial Intelligence, 2021....
we propose one-class graph neural network (OCGNN), a one-class classification framework for graph anomaly detection. OCGNN is designed to combine the powerful representation ability of graph neural networks along with the classical one-class objective. Compared with other baselines, OCGNN achieves ...
Code implementation for : Graph Neural Network-Based Anomaly Detection in Multivariate Time Series(AAAI'21) Installation Requirements Python >= 3.6 cuda == 10.2 Pytorch==1.5.1 PyG: torch-geometric==1.5.0 Install packages # run after installing correct Pytorch package bash install.sh Quick Start...
Xiao et al. [11] proposed a graph embedding approach to perform anomaly detection on network flows. The authors first converted the network flows into a first-order and secondorder graph. The first-order graph learns the latent features from the perspective of a single host by using its IP ...
回顾了异常检测(Anomaly Detection)、多元时间序列数据模型 (models for multivariate time series data)、图神经网络(Graph neural network)的研究相关工作,并指出其不足。 2.1 异常检测(Anomaly Detection) 目的是检测出偏离大部分数据的异常样本,经典方法包括基于密度的研究方法、基于线性模型的研究方法、基于距离的研究...
competitive graph neural network (CGNN)-based fraud detection system (eFraudCom) 一种基于竞争图神经网络(CGNN)的欺诈检测系统(eFraudCom),以检测电子商务平台上的欺诈行为。CGNN是一个基于GCN的GAE系统。eFraudCom系统由数据处理器和欺诈检测器组成。具体而言,在数据处理器中,对代表性的正态数据进行采样,并...