| Title |
A Neural Network Cost Model for Apartment Housing Projects in the Initial Stage |
| Keywords |
Neural Network ; Regression ; Error Back-Propagation Algorithm |
| Abstract |
The purpose of this research is to develope a prototype program model that builders are able to forecast construction costs in the first stage efficiently. This research is presenting the new side of estimating the construction cost by presenting the existing regression analysis model and the supplied neural network model and by comparing and verify two models to accomplish the estimating of first working expenses reasonably in the design and planning stage of apartment projects
The significance and outcome of this research is the same as: ⑴ Private construction companies that have used mutually different details-system quantity calculation methods could develop cost program model that is coincided with conditions in each construction companies. ⑵ We suggested the adequately applicable way in the construction of the neural network model to consider design and environmental factors that influence construction costs. |