Abstract
Important factors in the decision to bid process are identified. A rational and optimal model of decision making for the decision to bid, which depicts the relationships between the important factors and the decision to bid options, is developed. Parametric and non-parametric correlations are applied to the model to aid in understanding and demonstrating the linear approximation of the relationships in the model. The correlations satisfy the expected characteristics of a rational and optimal model of the decision domain. Regression models are employed to automate the rational and optimal model. The results indicate that to achieve an acceptable accuracy of prediction a tool that can capture non-linear relationships, in addition to linear relationships, needs to be applied to the model. Neural networks approach is such a tool and is utilized to automate the model.
| Original language | English |
|---|---|
| Title of host publication | Proceedings of the RICS Construction and Building Research Conference |
| Editors | John Kelly, Kirsty Hunter |
| Place of Publication | London |
| Publisher | RICS Foundation |
| Pages | 354-366 |
| Number of pages | 13 |
| Volume | 1 |
| ISBN (Print) | 1-84219-067-9 |
| Publication status | Published - 2001 |
Keywords
- bidding, modelling, decision support system, neural networks