A non-parametric outlier mining algorithm for detecting anomalies in construction equipment database

H. Fan, H. Kim, S. Abourizk

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

Abstract

Equipment management is one of the factors most critical to the success of construction projects. Being in charge of a large fleet, equipment managers for large contractors need to identify problems in equipment usage, repair, and maintenance based on large amounts of daily operational data. This paper presents a naïve outlier mining algorithm for the automatic screening and sorting of data sets based on the ways in which individual records deviate from their main bodies of data groups. The proposed outlier definition and outlier mining algorithm demonstrate better performance over the current well-known algorithms for real-world datasets. Application of the new algorithm to the proposed equipment analysis system is expected to provide reliable, quick solutions to problem detection in equipment databases.

Original languageEnglish
Title of host publicationProceedings - 33rd CSCE Annual Conference 2005
Subtitle of host publication6th Construction Specialty Conference
PagesCT-154-1-CT-154-10
Publication statusPublished - 2005 Dec 1
Event33rd CSCE Annual Conference 2005 - Toronto, ON, Canada
Duration: 2005 Jun 22005 Jun 4

Publication series

NameProceedings, Annual Conference - Canadian Society for Civil Engineering
Volume2005

Other

Other33rd CSCE Annual Conference 2005
CountryCanada
CityToronto, ON
Period05/6/205/6/4

Fingerprint

Construction equipment
Sorting
Contractors
Screening
Managers
Repair

All Science Journal Classification (ASJC) codes

  • Engineering(all)

Cite this

Fan, H., Kim, H., & Abourizk, S. (2005). A non-parametric outlier mining algorithm for detecting anomalies in construction equipment database. In Proceedings - 33rd CSCE Annual Conference 2005: 6th Construction Specialty Conference (pp. CT-154-1-CT-154-10). (Proceedings, Annual Conference - Canadian Society for Civil Engineering; Vol. 2005).
Fan, H. ; Kim, H. ; Abourizk, S. / A non-parametric outlier mining algorithm for detecting anomalies in construction equipment database. Proceedings - 33rd CSCE Annual Conference 2005: 6th Construction Specialty Conference. 2005. pp. CT-154-1-CT-154-10 (Proceedings, Annual Conference - Canadian Society for Civil Engineering).
@inproceedings{f11c9d0d789744cb952e7f78de093906,
title = "A non-parametric outlier mining algorithm for detecting anomalies in construction equipment database",
abstract = "Equipment management is one of the factors most critical to the success of construction projects. Being in charge of a large fleet, equipment managers for large contractors need to identify problems in equipment usage, repair, and maintenance based on large amounts of daily operational data. This paper presents a na{\"i}ve outlier mining algorithm for the automatic screening and sorting of data sets based on the ways in which individual records deviate from their main bodies of data groups. The proposed outlier definition and outlier mining algorithm demonstrate better performance over the current well-known algorithms for real-world datasets. Application of the new algorithm to the proposed equipment analysis system is expected to provide reliable, quick solutions to problem detection in equipment databases.",
author = "H. Fan and H. Kim and S. Abourizk",
year = "2005",
month = "12",
day = "1",
language = "English",
isbn = "1894662105",
series = "Proceedings, Annual Conference - Canadian Society for Civil Engineering",
pages = "CT--154--1--CT--154--10",
booktitle = "Proceedings - 33rd CSCE Annual Conference 2005",

}

Fan, H, Kim, H & Abourizk, S 2005, A non-parametric outlier mining algorithm for detecting anomalies in construction equipment database. in Proceedings - 33rd CSCE Annual Conference 2005: 6th Construction Specialty Conference. Proceedings, Annual Conference - Canadian Society for Civil Engineering, vol. 2005, pp. CT-154-1-CT-154-10, 33rd CSCE Annual Conference 2005, Toronto, ON, Canada, 05/6/2.

A non-parametric outlier mining algorithm for detecting anomalies in construction equipment database. / Fan, H.; Kim, H.; Abourizk, S.

Proceedings - 33rd CSCE Annual Conference 2005: 6th Construction Specialty Conference. 2005. p. CT-154-1-CT-154-10 (Proceedings, Annual Conference - Canadian Society for Civil Engineering; Vol. 2005).

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

TY - GEN

T1 - A non-parametric outlier mining algorithm for detecting anomalies in construction equipment database

AU - Fan, H.

AU - Kim, H.

AU - Abourizk, S.

PY - 2005/12/1

Y1 - 2005/12/1

N2 - Equipment management is one of the factors most critical to the success of construction projects. Being in charge of a large fleet, equipment managers for large contractors need to identify problems in equipment usage, repair, and maintenance based on large amounts of daily operational data. This paper presents a naïve outlier mining algorithm for the automatic screening and sorting of data sets based on the ways in which individual records deviate from their main bodies of data groups. The proposed outlier definition and outlier mining algorithm demonstrate better performance over the current well-known algorithms for real-world datasets. Application of the new algorithm to the proposed equipment analysis system is expected to provide reliable, quick solutions to problem detection in equipment databases.

AB - Equipment management is one of the factors most critical to the success of construction projects. Being in charge of a large fleet, equipment managers for large contractors need to identify problems in equipment usage, repair, and maintenance based on large amounts of daily operational data. This paper presents a naïve outlier mining algorithm for the automatic screening and sorting of data sets based on the ways in which individual records deviate from their main bodies of data groups. The proposed outlier definition and outlier mining algorithm demonstrate better performance over the current well-known algorithms for real-world datasets. Application of the new algorithm to the proposed equipment analysis system is expected to provide reliable, quick solutions to problem detection in equipment databases.

UR - http://www.scopus.com/inward/record.url?scp=33748956818&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33748956818&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:33748956818

SN - 1894662105

SN - 9781894662109

T3 - Proceedings, Annual Conference - Canadian Society for Civil Engineering

SP - CT-154-1-CT-154-10

BT - Proceedings - 33rd CSCE Annual Conference 2005

ER -

Fan H, Kim H, Abourizk S. A non-parametric outlier mining algorithm for detecting anomalies in construction equipment database. In Proceedings - 33rd CSCE Annual Conference 2005: 6th Construction Specialty Conference. 2005. p. CT-154-1-CT-154-10. (Proceedings, Annual Conference - Canadian Society for Civil Engineering).