Conceptual Cost Estimating: A Hybrid Neural-Expert System Approach

by Guruprasad N. Rao, Univ of Illinois at Urbana-Champaign, Urbana, United States,
Francois Grobler, Univ of Illinois at Urbana-Champaign, Urbana, United States,
Simon Kim, Univ of Illinois at Urbana-Champaign, Urbana, United States,



Document Type: Proceeding Paper

Part of: Computing in Civil and Building Engineering

Abstract: This paper presents a hybrid neural-expert system approach for developing conceptual cost estimates based on historical data of construction projects. The proposed hybrid neural-expert system approach has numerous advantages over other methods in that it can: consider a large number of project parameters (or factors); self-organize, that is, learn, from new information; function despite incomplete data input; and, explain the logic behind its conclusions. The aforementioned advantages of a hybrid neural-expert system is not only valid for comparison with other conventional conceptual cost estimating approaches, but also with systems developed using either expert systems or neural network techniques as stand-along systems.

Subject Headings: Hybrid methods | Expert systems | Construction costs | Neural networks | Parameters (statistics) | Information management | Comparative studies

Services: Buy this book/Buy this article

 

Return to search