Please use this identifier to cite or link to this item:
|Title:||Neural network method of estimating construction technology acceptability|
|Authors:||Chao, Li-Chung |
Skibniewski, Miroslaw J.
|Source:||Chao, Li-Chung, Skibniewski, Miroslaw J. (1995). Neural network method of estimating construction technology acceptability. Journal of Construction Engineering and Management 121 (1) : 130-142. ScholarBank@NUS Repository. https://doi.org/10.1061/(ASCE)0733-9364(1995)121:1(130)|
|Abstract:||A neural network (NN) based approach is proposed for predicting the adoption potential or acceptability of a new construction technology. The acceptability of a technology for a target operation is defined as the proportion of users that choose to use the technology in comparison to a conventional (base) technology. All existing alternative technologies for the considered operation are collected as samples for study. The performance characteristics of each sample technology are stored in a vector comprising eigenvalues determined by using the analytical hierarchy process (AHP) method, and its acceptability is determined using a poll. The obtained performance-acceptability pairs are used to train a neural network using the back-propagation algorithm. The trained network can then be used to predict the acceptability of a new technology in question, given its performance attributes. Possible information sources for training set construction and possible applications of the approach are discussed. An example estimate of the adoption prospect of a new concrete distribution system for concrete placement on a mid-rise building project is provided. Tests of the presented NN approach with simulated data show a promising result, especially when the poll size used is sufficiently large.|
|Source Title:||Journal of Construction Engineering and Management|
|Appears in Collections:||Staff Publications|
Show full item record
Files in This Item:
There are no files associated with this item.
checked on Dec 6, 2017
WEB OF SCIENCETM
checked on Nov 18, 2017
checked on Dec 10, 2017
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.