Anomalous Edge Detection in Edge Exchangeable Social Network Models
Preprint
Abstract
This paper studies detecting anomalous edges in directed graphs that model social networks. We exploit edge exchangeability as a criterion for distinguishing anomalous edges from normal edges. Then we present an anomaly detector based on conformal prediction theory; this detector has a guaranteed upper bound for false positive rate. In numerical experiments, we show that the proposed algorithm achieves superior performance to baseline methods.Keywords
All Related Versions
This publication has 10 references indexed in Scilit:
- Conformal prediction: A unified review of theory and new challengesBernoulli, 2023
- Algorithmic Learning in a Random World2022
- Edge Exchangeable Models for Interaction NetworksJournal of the American Statistical Association, 2018
- Probabilistic Foundations of Statistical Network Analysis2018
- Sparse Graphs Using Exchangeable Random MeasuresJournal of the Royal Statistical Society Series B: Statistical Methodology, 2017
- On Edge Exchangeable Random GraphsJournal of Statistical Physics, 2017
- Patterns and anomalies in k-cores of real-world graphs with applicationsKnowledge and Information Systems, 2017
- Anomaly detection in dynamic networks: a surveyWIREs Computational Statistics, 2015
- Graph based anomaly detection and description: a surveyData Mining and Knowledge Discovery, 2014
- Inductive Conformal Prediction: Theory and Application to Neural Networks2008