DocumentCode :
2505458
Title :
Anomalous subgraph detection via Sparse Principal Component Analysis
Author :
Singh, Navraj ; Miller, Benjamin A. ; Bliss, Nadya T. ; Wolfe, Patrick J.
Author_Institution :
Dept. of Electr. Eng., Stanford Univ., Stanford, CA, USA
fYear :
2011
fDate :
28-30 June 2011
Firstpage :
485
Lastpage :
488
Abstract :
Network datasets have become ubiquitous in many fields of study in recent years. In this paper we investigate a problem with applicability to a wide variety of domains - detecting small, anomalous subgraphs in a background graph. We characterize the anomaly in a subgraph via the well-known notion of network modularity, and we show that the optimization problem formulation resulting from our setup is very similar to a recently introduced technique in statistics called Sparse Principal Component Analysis (Sparse PCA), which is an extension of the classical PCA algorithm. The exact version of our problem formulation is a hard combinatorial optimization problem, so we consider a recently introduced semidefinite programming relaxation of the Sparse PCA problem. We show via results on simulated data that the technique is very promising.
Keywords :
graph theory; mathematical programming; network theory (graphs); principal component analysis; background graph; hard combinatorial optimization problem; network dataset; network modularity; optimization problem formulation; semidefinite programming relaxation; small anomalous subgraph detection; sparse principal component analysis; statistics; Approximation algorithms; Communities; Covariance matrix; Noise; Optimization; Principal component analysis; Programming; Anomaly detection; community detection; graph analysis; semidefinite programming; sparse principal component analysis;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Statistical Signal Processing Workshop (SSP), 2011 IEEE
Conference_Location :
Nice
ISSN :
pending
Print_ISBN :
978-1-4577-0569-4
Type :
conf
DOI :
10.1109/SSP.2011.5967738
Filename :
5967738
Link To Document :
بازگشت