Back to Search Start Over

Automatic manpower allocation for public construction projects using a rough set enhanced neural network

Authors :
Chen, Jieh-Haur
Yang, Li-Ren
Wang, Jui-Pin
Lin, Shang-I
Cheng, Jiun-Yao
Lee, Meng-Hsueh
Chen, Chih-Lin
Source :
Canadian Journal of Civil Engineering. August, 2021, Vol. 48 Issue 8, p1020, 6 p.
Publication Year :
2021

Abstract

Accurate estimates of manpower are still heavily dependent on well-experienced personnel. The objectives of this study are to prove the feasibility of using rough set theory to classify and weigh the impact attributes, and to develop a model to assess the total quantities of labor needed for a construction project using a rough set enhanced artificial neural network (ANN). Experts suggest 14 attributes that influence the estimation of on-site manpower for construction projects. After trimming and analyzing the basic data, the rough set approach is used to classify and weigh the attributes into three levels of impact based on their frequency. A rough set enhanced ANN is accordingly developed that yields an accuracy rate of 91.903%, higher than that of a regular ANN. A practical and effective prediction model benefits personnel having to estimate on-site manpower needs for construction projects. Key words: prediction, manpower, construction project, rough set, artificial neural network (ANN). Les estimations precises de la main-d'oeuvre dependent encore fortement d'un personnel experimente. Les objectifs de cette etude sont de prouver la faisabilite d'utiliser la theorie des ensembles bruts pour classer et ponderer les attributs d'impact, et de developper un modele pour evaluer les quantites totales de travail necessaires pour un projet de construction utilisant un reseau de neurones artificiels (RNA) ameliore. Les experts suggerent 14 attributs qui influent sur l'estimation de la main-d'oeuvre sur place pour les projets de construction. Apres avoir reduit et analyse les donnees de base, l'approche des ensembles bruts est utilisee pour classer et ponderer les attributs en trois niveaux d'impact en fonction de leur frequence. Un RNA ameliore par un ensemble brut est donc developpe qui donne un taux de precision de 91,903 %, superieur a celui d'un RNA regulier. Un modele de prevision pratique et efficace est avantageux pour le personnel qui doit estimer les besoins en main-d'oeuvre sur place pour les projets de construction. [Traduit par la Redaction] Mots-cles : prevision, main-d'oeuvre, projet de construction, ensemble brut, reseau de neurones artificiels (RNA).<br />Introduction Manpower is always one of the most important keys to the successful completion of a construction project. It comprises one of the largest proportions of most project budgets and [...]

Details

Language :
English
ISSN :
03151468
Volume :
48
Issue :
8
Database :
Gale General OneFile
Journal :
Canadian Journal of Civil Engineering
Publication Type :
Academic Journal
Accession number :
edsgcl.671460642
Full Text :
https://doi.org/10.1139/cjce-2019-0561