Open Access System for Information Sharing

Login Library

 

Article
Cited 0 time in webofscience Cited 0 time in scopus
Metadata Downloads

Learning method for knowledge retention in CBR cost models

Title
Learning method for knowledge retention in CBR cost models
Authors
Ji, Sae-Hyun안요섭LEE, EUL BUMKIM, YONGGU
POSTECH Authors
LEE, EUL BUM
Date Issued
Dec-2018
Publisher
ELSEVIER SCIENCE BV
Abstract
The case-based reasoning methodology fundamentally relies on historical cases to solve new problems. Supplementing insufficient data by the reproduction of appropriate values can mitigate the potential negative effects on the solutions resulting from sudden changes. However, CBR researchers have rarely examined this issue. To address this challenge, this research proposes a learning method for knowledge retention based on CBR by applying a data-mining approach to manage missing dataset values. A case study on a 164-apartment project was conducted to compare the estimation accuracy of the suggested learning method to that of past research with the same experiment conditions. The learning method with the CBR model achieved higher accuracy of the overall cost estimation and higher stability compared with the previous model. This research shows how cases can be generated and retained as learned cases to overcome the difficulties of continuous updates in a wide range of construction projects, as well as why the case bases need to be continuously updated. The research outcomes could support work related to cost estimation for decision makers ranging from beginners to experts in both academia and industry.
URI
http://oasis.postech.ac.kr/handle/2014.oak/93976
DOI
10.1016/j.autcon.2018.08.019
ISSN
0926-5805
Article Type
Article
Citation
AUTOMATION IN CONSTRUCTION, vol. 96, page. 65 - 74, 2018-12
Files in This Item:
There are no files associated with this item.

qr_code

  • mendeley

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.

Related Researcher

Researcher

 LEE, EUL BUM
엔지니어링 대학원
Read more

Altmetric

Views & Downloads

Browse