Building cost prediction at early design stage : artificial neural network approaches

Pao Yue-kong Library Electronic Theses Database

Building cost prediction at early design stage : artificial neural network approaches

 

Author: Yip, Kin-wa
Title: Building cost prediction at early design stage : artificial neural network approaches
Degree: M.Sc.
Year: 1997
Subject: Building -- Estimates
Neural networks (Computer science)
Hong Kong Polytechnic University -- Dissertations
Department: Multi-disciplinary Studies
Pages: ix, 88, [53] leaves : ill. ; 30 cm
Language: English
InnoPac Record: http://library.polyu.edu.hk/record=b1405409
URI: http://theses.lib.polyu.edu.hk/handle/200/4829
Abstract: One of the major professional services that quantity surveyors provide to their clients is to estimate or predict the construction costs of buildings at various stages of development projects. The cost predictions are crucial in studying the viability of the projects and in evaluating development alternatives. The nature of building cost prediction, and in particular that at early design stage, is such that qualitative judgments play an important part. Moreover, too much reliance on personal judgement may sometimes render the results subjective and inconsistent. Traditional computing technologies do not offer much in dealing with qualitative matters. Moreover, with the advent of different artificial intelligence techniques and in particular artificial neural networks, it is anticipated the problem can be approached from a new perspective. It is plausible that the requisite professional experience and intuition for cost prediction can be captured using some soft computing techniques and retrieved by any persons, thereby enhancing the efficiency and consistency of the work. A prototype for predicting the costs of lift installation was developed with regard to the above hypothesis. The prototype comprises two major components : an expert system and an artificial neural network. The expert system in the main serves to provide an user interface and to perform some fuzzification and other necessary pre- and post-processing calculations on the data in connection with the cost forecast. The forecast is done by the neural network, the network being set up using some genetic algorithms for optimum performance. The result is positive in general. Predicting ability of the prototype is within the normally acceptable standard in the construction industry. It is also indicated that the prototype is distinctly superior to the second generation of cost modelling based on regression analysis. In general, the study has demonstrated the capability of the artificial intelligence and other computing techniques in enhancing the work of the quantity surveyors. The project concludes to invoke further study in a number of directions, in particular, the extension of the approach for other building elements and the application of fuzzy logic in the problem domain.

Files in this item

Files Size Format
b14054097.pdf 4.953Mb PDF
Copyright Undertaking
As a bona fide Library user, I declare that:
  1. I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
  2. I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
  3. I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.

     

Quick Search

Browse

More Information