Sewer Condition Prediction and Analysis of Explanatory Factors

Sewer Condition Prediction and Analysis of Explanatory Factors

Sewer condition is commonly assessed using closed-circuit television (CCTV) inspections. In this paper, we combine inspection results, pipe attributes, network data, and data on pipe environment to predict pipe condition and to discover which factors affect it. We apply the random forest algorithm t...

Full description

Bibliographic Details
Main Authors: Tuija Laakso, Teemu Kokkonen, Ilkka Mellin, Riku Vahala
Format: Article
Language:English
Published: MDPI AG 2018-09-01
Series:Water
Subjects:
Boruta algorithm
logistic regression
partial dependence plot
random forest
sewer condition
variable selection
Online Access:http://www.mdpi.com/2073-4441/10/9/1239
id doaj-853be28b44f240a49bf470e3d01194ce
record_format Article
spelling doaj-853be28b44f240a49bf470e3d01194ce2020-11-25T00:46:08ZengMDPI AGWater2073-44412018-09-01109123910.3390/w10091239w10091239Sewer Condition Prediction and Analysis of Explanatory FactorsTuija Laakso0Teemu Kokkonen1Ilkka Mellin2Riku Vahala3Department of Built Environment, Aalto University, P.O. Box 15200, 00076 Aalto, FinlandDepartment of Built Environment, Aalto University, P.O. Box 15200, 00076 Aalto, FinlandDepartment of Mathematics and Systems Analysis, Aalto University, P.O. Box 11100, 00076 Aalto, FinlandDepartment of Built Environment, Aalto University, P.O. Box 15200, 00076 Aalto, FinlandSewer condition is commonly assessed using closed-circuit television (CCTV) inspections. In this paper, we combine inspection results, pipe attributes, network data, and data on pipe environment to predict pipe condition and to discover which factors affect it. We apply the random forest algorithm to model pipe condition and assess the variable importance using the Boruta algorithm. We analyse the impact of predictor variables on poor condition using partial dependence plots, which are a valuable technique for this purpose. The results can be used in screening pipes for future inspections and provide insight into the dynamics between predictor variables and poor condition.http://www.mdpi.com/2073-4441/10/9/1239Boruta algorithmlogistic regressionpartial dependence plotrandom forestsewer conditionvariable selection
collection DOAJ
language English
format Article
sources DOAJ
author Tuija Laakso
Teemu Kokkonen
Ilkka Mellin
Riku Vahala
spellingShingle Tuija Laakso
Teemu Kokkonen
Ilkka Mellin
Riku Vahala
Sewer Condition Prediction and Analysis of Explanatory Factors
Water
Boruta algorithm
logistic regression
partial dependence plot
random forest
sewer condition
variable selection
author_facet Tuija Laakso
Teemu Kokkonen
Ilkka Mellin
Riku Vahala
author_sort Tuija Laakso
title Sewer Condition Prediction and Analysis of Explanatory Factors
title_short Sewer Condition Prediction and Analysis of Explanatory Factors
title_full Sewer Condition Prediction and Analysis of Explanatory Factors
title_fullStr Sewer Condition Prediction and Analysis of Explanatory Factors
title_full_unstemmed Sewer Condition Prediction and Analysis of Explanatory Factors
title_sort sewer condition prediction and analysis of explanatory factors
publisher MDPI AG
series Water
issn 2073-4441
publishDate 2018-09-01
description Sewer condition is commonly assessed using closed-circuit television (CCTV) inspections. In this paper, we combine inspection results, pipe attributes, network data, and data on pipe environment to predict pipe condition and to discover which factors affect it. We apply the random forest algorithm to model pipe condition and assess the variable importance using the Boruta algorithm. We analyse the impact of predictor variables on poor condition using partial dependence plots, which are a valuable technique for this purpose. The results can be used in screening pipes for future inspections and provide insight into the dynamics between predictor variables and poor condition.
topic Boruta algorithm
logistic regression
partial dependence plot
random forest
sewer condition
variable selection
url http://www.mdpi.com/2073-4441/10/9/1239
work_keys_str_mv AT tuijalaakso sewerconditionpredictionandanalysisofexplanatoryfactors
AT teemukokkonen sewerconditionpredictionandanalysisofexplanatoryfactors
AT ilkkamellin sewerconditionpredictionandanalysisofexplanatoryfactors
AT rikuvahala sewerconditionpredictionandanalysisofexplanatoryfactors
_version_ 1725266708170539008

Similar Items