A causal approach for mining interesting anomalies

Sakshi Babbar, Didi Surian, Sanjay Chawla

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

We propose a novel approach which combines the use of Bayesian network and probabilistic association rules to discover and explain anomalies in data. The Bayesian network allows us to organize information in order to capture both correlation and causality in the feature space, while the probabilistic association rules have a structure similar to association mining rules. In particular, we focus on two types of rules: (i) low support & high confidence and, (ii) high support & low confidence. New data points which satisfy either one of the two rules conditioned on the Bayesian network are the candidate anomalies. We perform extensive experiments on well-known benchmark data sets and demonstrate that our approach is able to identify anomalies in high precision and recall. Moreover, our approach can be used to discover contextual information from the mined anomalies, which other techniques often fail to do so.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Pages226-232
Number of pages7
Volume7884 LNAI
DOIs
Publication statusPublished - 2013
Externally publishedYes
Event26th Canadian Conference on Artificial Intelligence, Canadian AI 2013 - Regina, SK
Duration: 28 May 201331 May 2013

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume7884 LNAI
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other26th Canadian Conference on Artificial Intelligence, Canadian AI 2013
CityRegina, SK
Period28/5/1331/5/13

Fingerprint

Bayesian networks
Anomaly
Mining
Bayesian Networks
Association rules
Association Rules
Confidence
Association Rule Mining
Causality
Feature Space
Benchmark
Experiments
Demonstrate
Experiment

Keywords

  • anomaly
  • Bayesian network
  • causality
  • probabilistic association rules

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Babbar, S., Surian, D., & Chawla, S. (2013). A causal approach for mining interesting anomalies. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 7884 LNAI, pp. 226-232). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7884 LNAI). https://doi.org/10.1007/978-3-642-38457-8_20

A causal approach for mining interesting anomalies. / Babbar, Sakshi; Surian, Didi; Chawla, Sanjay.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7884 LNAI 2013. p. 226-232 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 7884 LNAI).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Babbar, S, Surian, D & Chawla, S 2013, A causal approach for mining interesting anomalies. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 7884 LNAI, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 7884 LNAI, pp. 226-232, 26th Canadian Conference on Artificial Intelligence, Canadian AI 2013, Regina, SK, 28/5/13. https://doi.org/10.1007/978-3-642-38457-8_20
Babbar S, Surian D, Chawla S. A causal approach for mining interesting anomalies. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7884 LNAI. 2013. p. 226-232. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-642-38457-8_20
Babbar, Sakshi ; Surian, Didi ; Chawla, Sanjay. / A causal approach for mining interesting anomalies. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 7884 LNAI 2013. pp. 226-232 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{59de19b96b894bc7820e5cb26e550e45,
title = "A causal approach for mining interesting anomalies",
abstract = "We propose a novel approach which combines the use of Bayesian network and probabilistic association rules to discover and explain anomalies in data. The Bayesian network allows us to organize information in order to capture both correlation and causality in the feature space, while the probabilistic association rules have a structure similar to association mining rules. In particular, we focus on two types of rules: (i) low support & high confidence and, (ii) high support & low confidence. New data points which satisfy either one of the two rules conditioned on the Bayesian network are the candidate anomalies. We perform extensive experiments on well-known benchmark data sets and demonstrate that our approach is able to identify anomalies in high precision and recall. Moreover, our approach can be used to discover contextual information from the mined anomalies, which other techniques often fail to do so.",
keywords = "anomaly, Bayesian network, causality, probabilistic association rules",
author = "Sakshi Babbar and Didi Surian and Sanjay Chawla",
year = "2013",
doi = "10.1007/978-3-642-38457-8_20",
language = "English",
isbn = "9783642384561",
volume = "7884 LNAI",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "226--232",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - A causal approach for mining interesting anomalies

AU - Babbar, Sakshi

AU - Surian, Didi

AU - Chawla, Sanjay

PY - 2013

Y1 - 2013

N2 - We propose a novel approach which combines the use of Bayesian network and probabilistic association rules to discover and explain anomalies in data. The Bayesian network allows us to organize information in order to capture both correlation and causality in the feature space, while the probabilistic association rules have a structure similar to association mining rules. In particular, we focus on two types of rules: (i) low support & high confidence and, (ii) high support & low confidence. New data points which satisfy either one of the two rules conditioned on the Bayesian network are the candidate anomalies. We perform extensive experiments on well-known benchmark data sets and demonstrate that our approach is able to identify anomalies in high precision and recall. Moreover, our approach can be used to discover contextual information from the mined anomalies, which other techniques often fail to do so.

AB - We propose a novel approach which combines the use of Bayesian network and probabilistic association rules to discover and explain anomalies in data. The Bayesian network allows us to organize information in order to capture both correlation and causality in the feature space, while the probabilistic association rules have a structure similar to association mining rules. In particular, we focus on two types of rules: (i) low support & high confidence and, (ii) high support & low confidence. New data points which satisfy either one of the two rules conditioned on the Bayesian network are the candidate anomalies. We perform extensive experiments on well-known benchmark data sets and demonstrate that our approach is able to identify anomalies in high precision and recall. Moreover, our approach can be used to discover contextual information from the mined anomalies, which other techniques often fail to do so.

KW - anomaly

KW - Bayesian network

KW - causality

KW - probabilistic association rules

UR - http://www.scopus.com/inward/record.url?scp=84884497790&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84884497790&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-38457-8_20

DO - 10.1007/978-3-642-38457-8_20

M3 - Conference contribution

AN - SCOPUS:84884497790

SN - 9783642384561

VL - 7884 LNAI

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 226

EP - 232

BT - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

ER -